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

Cost Functions

by

Elmer G. Wiens

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

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

N. Generalized CES-Diewert (Generalized Leontief) Cost Function

This web page replicates the procedures used to estimate the parameters of a Diewert cost function to approximate a CES cost function. To determine the Diewert cost function's efficacy in estimating varying elasticities of substitution among pairs of inputs, on this web page we approximate a Generalized CES cost function with the Diewert cost function.

Unlike the CES cost function, the Generalized CES cost function is not necessarily homothetic, while the Diewert cost function is homothetic by construction. Moreover, the Generalized CES's elasticity of scale is a function of its factor inputs, i.e. its elasticity of scale, εLKM, varies with factor prices and with output.

Consequently, when we estimate the Diewert cost function, we need to estimate the parameter nu1 of its returns to scale function separately, as we did when we approximated the CES production function by the Diewert production function.

The three factor Diewert (Generalized Leontief) (total) cost function is:

C(q;wL,wK,wM) = h(q) * c(wL,wK,wM)         (**)

where the returns to scale function is:

h(q) = q^(1/nu1)

a continuous, increasing function of q (q >= 1), with h(0) = 0 and h(1) = 1, where nu1 is a measure of the returns to scale,

and the unit cost function is:

c(wL,wK,wM) = cLL * wL + cKK * wK + cMM * wM + (dLK+dKL) * (wL*wK)^(1/2) + (dLM +dML) * (wL*wM)^(1/2) + (dKM+dMK) * (wK*wM)^(1/2)

linear in its parameters cLL, cKK, cMM, dLK, dKL, dLM, dML, dKM, and dMK.

We shall use two methods to obtain estimates of the parameters:

          À1: Estimate the parameters the cost function by way of its factor share functions — requires data relating output quantities to (cost minimizing) factor inputs, and input prices,

          À2: Estimate the parameters of the cost function directly — requires data relating output quantities to total (minimum) cost and input prices.

We shall use the three factor Generalized CES production function to generate the required cost data, yielding a data set relating output, q, and factor prices, wL, wK, and wM, [randomized about the base prices (wL*, wK*, wM*)]   to the Generalized CES cost minimizing factor inputs, L, K, and M.

For methods À1, and À2, we shall set nu1 = εLKM where q = 30 units of output, i.e. at the elasticity of scale of the Generalized CES cost function for q = 30.

To illustrate what can happen when nu1 is set incorrectly, we shall repeat methods À1, and À2 as methods À3, and À4, with nu1 = 1, i.e. under the assumption of constant returns to scale along the domain of q.

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.

Generalized CES Elasticity of Scale of Production:

εLKM = ε(L,K,M) = (nu / rho) * (alpha * rhoL * L^-rhoL + beta * rhoK * K^-rhoK + gamma * rhoM * M^-rhoM) / (alpha * L^-rhoL + beta * K^-rhoK + gamma * M^-rhoM).

See the Generalized CES production function.



À1:   Estimate the factor demand equations separately.

Assuming dLK = dKL, dLM = dML, and dKM = dMK, the Diewert unit cost function becomes:

c(wL,wK,wM) = cLL * wL + cKK * wK + cMM * wM + 2 * dLK * (wL*wK)^(1/2) + 2 * dLM * (wL*wM)^(1/2) + 2 * dKM * (wK*wM)^(1/2)

Taking the partial derivative of the cost function with respect to an input price, we get the factor demand function for that input:

∂C/∂wL = L(q; wL, wK, wM) = q^(1/nu1) * [cLL + dLK * (wK / wL)^(1/2) + dLM * (wM / wL)^(1/2)]

∂C/∂wK = K(q; wL, wK, wM) = q^(1/nu1) * [cKK + dKL * (wL / wK)^(1/2) + dKM * (wM / wK)^(1/2)]

∂C/∂wM = M(q; wL, wK, wM) = q^(1/nu1) * [cMM + dML * (wL / wM)^(1/2) + dMK * (wK / wM)^(1/2)]

Since it is likely that, as measured numerically, dLK != dKL, dLM != dML, and dKM != dMK, these distinctions are maintained in the factor demand equations.

After dividing each factor demand equation by q(1/nu1), we can estimate the parameters of these three linear in parameters equations using linear multiple regression.

Having obtained the unrestricted estimates of the coefficients of the factor demand functions, we shall compute the restricted least squares estimates with dLK = dKL, dLM = dML, and dKM = dMK.

The substituted values of the estimated parameters determine the Diewert cost function.

I. Stage 1. Generate cost data with the Generalized CES production function, yielding a data set relating output, q, and factor prices, wL, wK, and wM, [randomized about the base prices (wL*, wK*, wM*)]   to the Generalized CES cost minimizing factor inputs, L, K, and M. The base factor prices and the randomizing distribution are specified in the form below.

   Stage 2. Obtain the Diewert cost function by substituting the estimates (obtained via linear multiple regression) of the parameters of the three factor demand equations into the cost function.

II. The estimated coefficients of the Diewert cost function 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.

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.

Generalized CES elasticity of scale at q = 30:     εLKM = nu1 = 0.92

III. For these coefficients of the CES Generalized production function, I generated a sequence (displayed in the "Generalized CES-Diewert Cost Function" table) of factor prices, outputs, and the corresponding cost minimizing inputs. Then I used these data to estimate the coefficients of each factor share equation separately:

QR Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
cLL-0.1016750.004-25.368
dLK0.51630.003174.164
dLM0.5234520.003178.341
R2 = 0.9997 R2b = 0.9997 # obs = 31

QR Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
cKK-0.0096530.019-0.499
dKL0.479610.02518.859
dKM0.3798610.01919.929
R2 = 0.9746 R2b = 0.9728 # obs = 31

QR Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
cMM-0.1400670.029-4.843
dML0.5782470.03715.69
dMK0.3464510.02613.283
R2 = 0.9834 R2b = 0.9822 # obs = 31

The three estimated factor demand functions are:

∂C/∂wL = L(q; wL, wK, wM) = q^(1/0.92) * [-0.1017 + 0.5163 * (wK / wL)^(1/2) + 0.5235 * (wM / wL)^(1/2)]

∂C/∂wK = K(q; wL, wK, wM) = q^(1/0.92) * [-0.0097 + 0.4796 * (wL / wK)^(1/2) + 0.3799 * (wM / wK)^(1/2)]

∂C/∂wM = M(q; wL, wK, wM) = q^(1/0.92) * [-0.1401 + 0.5782 * (wL / wM)^(1/2) + 0.3465 * (wK / wM)^(1/2)]

As estimated, generally dLK != dKL, dLM != dML, and dKM != dMK: Young's Theorem doesn't hold without constraints across equations. Consequently, I use the sum of the cross parameters in the Diewert cost function.

    The estimated factor demand elasticities are obtained by:

εL,wL = ∂ln(L(q;wL,wK,wM))/∂ln(wL) = -.5 * (q^1/nu1) / L(q;wL,wK,wM)) * (dLK * (wK/wL)^1/2 + dLM * (wM/wL)^1/2),

εL,wK = ∂ln(L(q;wL,wK,wM))/∂ln(wK) = .5 * (q^1/nu1) / L(q;wL,wK,wM)) * dLK * (wK/wL)^1/2,

εL,wM = ∂ln(L(q;wL,wK,wM))/∂ln(wM) = .5 * (q^1/nu1) / L(q;wL,wK,wM)) * dLM * (wM/wL)^1/2,

εL,q = ∂ln(L(q;wL,wK,wM))/∂ln(q) = 1/nu1, etc.

      For example, with wL = 7, wK = 13, wM = 6, and q = 30:

εL,wLεL,wKεL,wMεL,q
εK,wLεK,wKεK,wMεK,q
εM,wLεM,wKεM,wMεM,q
-0.54680.32380.2231.087
0.2931-0.5080.21491.087
0.3140.2564-0.57041.087

IV. The Diewert cost function as obtained from the unrestricted factor demand equations is:

C(q;wL,wK,wM) = q^(1/nu1) * [cLL * wL + cKK * wK + cMM * wM + (dLK+dKL) * (wL*wK)^(1/2) + (dLM +dML) * (wL*wM)^(1/2) + (dKM+dMK) * (wK*wM)^(1/2)]

= q^(1/0.92) * [-0.10168 * wL + -0.00965 * wK + -0.14007 * wM + 0.99591 * (wL*wK)^(1/2) + 1.1017 * (wL*wM)^(1/2) + 0.72631 * (wK*wM)^(1/2)]
        (***)

V. The Restricted Factor Demand Equations.

Having obtained the unrestricted estimates of the coefficients of the factor demand functions, we compute the restricted least squares estimates with dLK = dKL, dLM = dML, and dKM = dMK.

QR Restricted Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
cLL-0.1035750.023433-4.42011
dLK0.5036770.01629630.90709
dLM0.544370.01410538.593551
cKK-0.0180240.015082-1.195068
dKL0.5036770.01629630.90709
dKM0.3659780.01203330.414695
cMM-0.1320590.017127-7.710364
dML0.544370.01410538.593551
dMK0.3659780.01203330.414695
R2 = 0.9973 R2b = 0.997 # obs = 93

dLK = dKL, dLM = dML, and dKM = dMK

The three estimated, restricted factor demand functions are:

∂C/∂wL = L(q; wL, wK, wM) = q^(1/0.92) * [-0.1036 + 0.5037 * (wK / wL)^(1/2) + 0.5444 * (wM / wL)^(1/2)]

∂C/∂wK = K(q; wL, wK, wM) = q^(1/0.92) * [-0.018 + 0.5037 * (wL / wK)^(1/2) + 0.366 * (wM / wK)^(1/2)]

∂C/∂wM = M(q; wL, wK, wM) = q^(1/0.92) * [-0.1321 + 0.5444 * (wL / wM)^(1/2) + 0.366 * (wK / wM)^(1/2)]

    The estimated factor demand elasticities are obtained by:

εL,wL = ∂ln(L(q;wL,wK,wM))/∂ln(wL) = -.5 * (q^1/nu1) / L(q;wL,wK,wM)) * (dLK * (wK/wL)^1/2 + dLM * (wM/wL)^1/2),

εL,wK = ∂ln(L(q;wL,wK,wM))/∂ln(wK) = .5 * (q^1/nu1) / L(q;wL,wK,wM)) * dLK * (wK/wL)^1/2,

εL,wM = ∂ln(L(q;wL,wK,wM))/∂ln(wM) = .5 * (q^1/nu1) / L(q;wL,wK,wM)) * dLM * (wM/wL)^1/2,

εL,wq = ∂ln(L(q;wL,wK,wM))/∂ln(q) = 1/nu, etc.

      For example, with wL = 7, wK = 13, wM = 6, and q = 30:

εL,wLεL,wKεL,wMεL,q
εK,wLεK,wKεK,wMεK,q
εM,wLεM,wKεM,wMεM,q
-0.54770.31580.23191.087
0.3079-0.5150.20711.087
0.29560.2708-0.56641.087

VI. The Diewert cost function as obtained from the restricted factor demand equations is:

C(q;wL,wK,wM) = q^(1/nu) * [cLL * wL + cKK * wK + cMM * wM + (dLK+dKL) * (wL*wK)^(1/2) + (dLM +dML) * (wL*wM)^(1/2) + (dKM+dMK) * (wK*wM)^(1/2)]

= q^(1/0.92) * [-0.10358 * wL + -0.01802 * wK + -0.13206 * wM + 1.00735 * (wL*wK)^(1/2) + 1.08874 * (wL*wM)^(1/2) + 0.73196 * (wK*wM)^(1/2)]
        (***)

VII. Note:
      1. C(q;wL,wK,wM) linear homogeneous in factor prices requires for any t > 0 that the unit cost function c(wL,wK,wM) obey:

c(t*wL,t*wK,t*wM) = t * c(wL,wK,wM).

        For example, with wL = 7, wK = 13, wM = 6, and t = 2:

2 * c(7, 13, 6) = 42.76 = 42.76 = c(14, 26,12).

      2. The matrix ∇2C of second order partial derivatives of the cost function C(q;wL,wK,wM) = h(q)*c(wL,wK,wM) is symmetric:

2C/∂q∂wL = h'(q) * ∂c/∂wL = ∂2C/∂wL∂q,

2C/∂q∂wK = h'(q) * ∂c/∂wK = ∂2C/∂wK∂q,

2C/∂q∂wM = h'(q) * ∂c/∂wM = ∂2C/∂wM∂q,

2c/∂wL∂wK = .25 * (dLK+dKL)/(wL*wK)^(1/2) = ∂2c/∂wK∂wL,

2c/∂wL∂wM = .25 * (dLM+dML)/(wL*wM)^(1/2) = ∂2c/∂wM∂wL,

2c/∂wK∂wM = .25 * (dKM+dMK)/(wL*wK)^(1/2) = ∂2c/∂wM∂wK, and

2c/∂wL∂wL = -.25((dLK+dKL)*wK^(1/2) + (dLM+dML)*wM^(1/2))/wL^(3/2),

2c/∂wK∂wK = -.25((dKL+dLK)*wL^(1/2) + (dKM+dMK)*wM^(1/2))/wK^(3/2),

2c/∂wM∂wM = -.25((dML+dLM)*wL^(1/2) + (dMK+dKM)*wK^(1/2))/wM^(3/2).

      3. The cost function C(q;wL,wK,wM) is a concave in factor prices if the Hessian matrix H = ∇2wwc(wL,wK,wM) of second order partial derivatives with respect to factor prices is negative semidefinite.   The values of the second order partial derivatives depend on factor prices. As an example, consider the case where wL = 7, wK = 13, and wM = 6:

H = ∇2wwc =
-0.085030.02640.042
0.0264-0.023780.02072
0.0420.02072-0.09389

The principal minors of H are H1 = -0.085027, H2 = 0.001325, and H3 = -0.

If these principal minors alternate in sign, starting with negative, with H3 <= 0, the matrix H is negative (semi)definite,
and the cost function C(q;wL,wK,wM) is a concave function in factor prices at wL = 7, wK = 13, and wM = 6.

The eigenvalues of H are e1 = -0.1317, e2 = -0.071, and e3 = -0.
H3 = e1 * e2 * e3 = -0.

VIII. The factor share functions are:

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

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

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

IX. Uzawa Partial Elasticities of Substitution:

uLK = C(q;wL,wK,wM) * ∂L(q;wL,wK,wM)/∂wK / (L(q;wL,wK,wM) * K(q;wL,wK,wM)),

uLM = C(q;wL,wK,wM) * ∂L(q;wL,wK,wM)/∂wM / (L(q;wL,wK,wM) * M(q;wL,wK,wM)),

uKL = C(q;wL,wK,wM) * ∂K(q;wL,wK,wM)/∂wL / (K(q;wL,wK,wM) * L(q;wL,wK,wM)),

uKM = C(q;wL,wK,wM) * ∂K(q;wL,wK,wM)/∂wM / (K(q;wL,wK,wM) * M(q;wL,wK,wM)),

uML = C(q;wL,wK,wM) * ∂M(q;wL,wK,wM)/∂wL / (M(q;wL,wK,wM) * L(q;wL,wK,wM)),

uMK = C(q;wL,wK,wM) * ∂M(q;wL,wK,wM)/∂wK / (M(q;wL,wK,wM) * M(q;wL,wK,wM)),

Imposing the dLK = dKL, dLM = dML, and dKM = dMK restrictions in the factor demand functions, we get (approximate?) equality of the cross partial elasticities of substitution:

uLK = uKL,    uLM = uML,  and    uKM = uMK.

as seen in the following table.

X. Table of Results: check that the estimated Diewert cost function and input amounts for a given level of output agree with the Generalized CES production function's minimized cost and inputs. Also, compare the Allen partial elasticities of substitution, sLK, sLM, and sMK, with the Uzawa partial elasticities of substitution, uLK, uLM, and uKM.

Generalized CES-Diewert Cost Function
À1: Estimate the Translog cost function using restricted factor shares
Parameters: rho = 0.17647, rhoL = 0.17647, rhoK = 0.11823, rhoM = 0.26339
 nu = 1
  —   Generalized CES Cost Data   —  
nu1 = 0.92
  —   Diewert Cost Data   —  
Factor SharesUzawa Elasticities2wwc(w)
obs #qwLwKwM LK MsLKsLMsKMεLKMcostest costest Lest Kest MsLsKsMuLKuLMuKLuKMuMLuMKe1e2e3
1188 13.027.02 24.2814.61 22.740.90.790.830.93544.08542.6524.2814.9421.930.3580.3580.2840.850.860.850.730.860.73 -0.111-0.063-0
2197.46 11.287.92 26.3116.97 21.080.90.790.840.934554.62552.3926.4317.1420.450.3570.350.2930.820.890.820.750.890.75 -0.108-0.064-0
3208.86 11.84.44 22.5216.15 33.040.890.790.830.922536.81537.4622.416.6832.010.3690.3660.2640.920.840.920.660.840.66 -0.172-0.065-0
4216.34 11.584.1 27.7815.18 32.830.890.790.830.92486.47487.6127.7715.6631.740.3610.3720.2670.90.810.90.710.810.71 -0.187-0.0840
5227.26 12.57.44 31.8118.46 25.870.90.790.830.929654.13653.3131.918.6625.330.3550.3570.2880.840.860.840.750.860.75 -0.115-0.064-0
6236.98 12.44.86 30.8917.4 34.180.890.790.830.921597.44598.3530.8717.7933.380.360.3690.2710.890.820.890.720.820.72 -0.156-0.0760
7245.72 11.626.64 37.9219.15 27.60.90.790.830.927622.71622.4537.9919.3627.130.3490.3610.2890.830.840.830.780.840.78 -0.145-0.0710
8256.46 11.947.04 37.8920.82 290.90.790.830.927697.52697.4238.0220.9528.640.3520.3590.2890.830.850.830.770.850.77 -0.128-0.0670
9268.02 12.567.92 36.6623.32 30.40.90.790.830.927827.61827.8536.8523.330.260.3570.3540.2890.830.880.830.740.880.74 -0.104-0.0610
10276.9 14.94.84 39.2718.81 42.490.890.790.830.915756.89758.7539.2919.1841.720.3570.3770.2660.90.780.90.730.780.73 -0.163-0.076-0
11287.82 11.685.86 36.7624.4 37.920.890.790.830.922794.59795.3136.7824.4437.920.3620.3590.2790.870.860.870.710.860.71 -0.126-0.07-0
12297.96 12.666.24 39.3524.78 38.990.890.790.830.921870.3871.1539.3924.8139.020.360.3610.280.870.850.870.720.850.72 -0.121-0.0670
13307 136 43.7924.14 40.130.890.790.830.92861.17862.0643.8224.240.110.3560.3650.2790.870.830.870.740.830.74 -0.132-0.071-0
14318 14.024.52 40.7123.55 52.150.890.790.830.913891.55892.7440.6323.8351.680.3640.3740.2620.920.80.920.70.80.7 -0.172-0.068-0
15328.92 11.786.42 40.4830.15 42.830.890.790.830.922991.2992.5440.5329.8643.490.3640.3540.2810.870.880.870.70.880.7 -0.113-0.064-0
16335 11.587.98 60.1527.67 32.970.90.790.830.926884.25885.4960.4127.5833.080.3410.3610.2980.790.850.790.830.850.83 -0.167-0.061-0
17346.18 14.484.04 51.1823.11 57.490.890.790.830.909883.12883.9751.1723.357.020.3580.3820.2610.910.760.910.740.760.74 -0.199-0.086-0
18357.36 12.627.42 52.3131.26 41.650.890.790.830.9221088.511090.3252.5730.8642.310.3550.3570.2880.840.860.840.750.860.75 -0.114-0.063-0
19365.48 147.9 66.6628.24 39.320.890.790.830.9211071.321072.866.6228.1239.750.340.3670.2930.810.820.810.820.820.82 -0.155-0.059-0
20375.02 12.587.38 68.5529.59 39.740.890.790.830.9211009.591011.1268.5729.440.250.340.3660.2940.810.830.810.820.830.82 -0.168-0.064-0
21388.36 12.424.9 48.5432.84 59.660.890.790.830.9141106.031105.6548.3432.5960.570.3650.3660.2680.90.840.90.690.840.69 -0.153-0.067-0
22395.4 14.424.36 65.0126.44 60.780.890.790.830.909997.32996.5164.8226.3661.10.3510.3810.2670.890.760.890.770.760.77 -0.191-0.094-0
23408.02 14.384.58 54.3631.28 67.690.890.790.830.9091195.731194.1754.1531.1368.180.3640.3750.2610.920.80.920.70.80.7 -0.17-0.0670
24415.3 13.944.14 67.8628.04 64.940.890.790.830.9081019.321017.7267.6427.8665.430.3520.3820.2660.90.760.90.760.760.76 -0.199-0.097-0
25425.84 11.584.14 62.0432.85 64.510.890.790.830.9111009.721008.2861.8532.4765.490.3580.3730.2690.890.80.890.730.80.73 -0.187-0.089-0
26437.64 13.744.84 60.1834.72 68.540.890.790.830.911268.541266.2259.9434.2869.690.3620.3720.2660.90.810.90.710.810.71 -0.158-0.07-0
27446.88 14.785.3 68.2533.81 66.010.890.790.830.911319.091316.5768.0233.3167.220.3550.3740.2710.890.790.890.740.790.74 -0.149-0.0740
28457.16 12.988 71.6341.4 51.450.890.790.830.9191461.851464.872.0640.3153.20.3520.3570.2910.830.860.830.770.860.77 -0.114-0.060
29466.94 11.264.84 63.8740.52 67.140.890.790.830.9121224.541222.9663.739.6269.180.3610.3650.2740.890.840.890.710.840.71 -0.155-0.0780
30475.86 14.784.1 7632.79 78.740.890.790.830.9051252.881247.7575.6532.3179.720.3550.3830.2620.910.750.910.750.750.75 -0.198-0.090
31487.6 12.46.06 68.9743.71 64.860.890.790.830.9141459.231458.5968.9542.4967.280.3590.3610.280.870.850.870.720.850.72 -0.125-0.069-0
AVE:336.96 12.865.85 49.0826.54 46.590.890.790.830.918933.49933.3949.0826.5446.590.3570.3660.2770.870.830.870.740.830.74-0.15-0.072-0




À2:   Estimate the Diewert cost function directly.

XI. If we have a data set relating the input (factor) prices, wL, wK, and wM, to the total (minimum) cost of producing output for varying levels output, but do not have data on the required levels of inputs, we can estimate the Diewert cost function directly. We will use the same sequence (displayed in the "Diewert Cost Function" table) of factor prices, outputs, and total (minimum) cost as used in À1. The estimated coefficients of the cost function will vary with the parameters nu, rho, rhoL, rhoK, rhoM, alpha, beta and gamma of the Generalized CES production function used to generate these data.

QR Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
cLL-0.1621280.097-1.674
cKK0.0543410.0580.943
cMM-0.0724090.093-0.775
2*dLK0.9880130.1446.853
2*dLM1.2502020.0913.934
2*dKM0.5328980.1124.756
R2 = 0.9996 R2b = 0.9996 # obs = 31

The Diewert cost function as obtained by direct estimation is:

C(q;wL,wK,wM) = q^(1/nu1) * [cLL * wL + cKK * wK + cMM * wM + 2 * dLK * (wL*wK)^(1/2) + 2 * dLM * (wL*wM)^(1/2) + 2 * dKM * (wK*wM)^(1/2)]

= q^(1/0.92) * [-0.16213 * wL + 0.05434 * wK + -0.07241 * wM + 0.98801 * (wL*wK)^(1/2) + 1.2502 * (wL*wM)^(1/2) + 0.5329 * (wK*wM)^(1/2)]
        (***)

XII. Its three derived factor demand functions are:

∂C/∂wL = L(q; wL, wK, wM) = q^(1/0.92) * [-0.1621 + 0.494 * (wK / wL)^(1/2) + 0.6251 * (wM / wL)^(1/2)]

∂C/∂wK = K(q; wL, wK, wM) = q^(1/0.92) * [0.0543 + 0.494 * (wL / wK)^(1/2) + 0.2664 * (wM / wK)^(1/2)]

∂C/∂wM = M(q; wL, wK, wM) = q^(1/0.92) * [-0.0724 + 0.6251 * (wL / wM)^(1/2) + 0.2664 * (wK / wM)^(1/2)]

    The derived factor demand elasticities are obtained by:

εL,wL = ∂ln(L(q;wL,wK,wM))/∂ln(wL) = -.5 * (q^1/nu1) / L(q;wL,wK,wM)) * (dLK * (wK/wL)^1/2 + dLM * (wM/wL)^1/2),

εL,wK = ∂ln(L(q;wL,wK,wM))/∂ln(wK) = .5 * (q^1/nu1) / L(q;wL,wK,wM)) * dLK * (wK/wL)^1/2,

εL,wM = ∂ln(L(q;wL,wK,wM))/∂ln(wM) = .5 * (q^1/nu1) / L(q;wL,wK,wM)) * dLM * (wM/wL)^1/2,

εL,q = ∂ln(L(q;wL,wK,wM))/∂ln(q) = 1/nu, etc.

      For example, with wL = 7, wK = 13, wM = 6, and q = 30:

εL,wLεL,wKεL,wMεL,q
εK,wLεK,wKεK,wMεK,q
εM,wLεM,wKεM,wMεM,q
-0.57440.30890.26551.087
0.3032-0.45460.15141.087
0.33930.1971-0.53641.087

XIII. Notes:
      1. As derived, dLK = dKL, dLM = dML, and dKM = dMK: Young's Theorem holds by construction.

      2. C(q;wL,wK,wM) linear homogeneous in factor prices requires for any t > 0 that the unit cost function c(wL,wK,wM) obey:

c(t*wL,t*wK,t*wM) = t * c(wL,wK,wM).

        For example, with wL = 7, wK = 13, wM = 6, and t = 2:

2 * c(7, 13, 6) = 42.74 = 42.74 = c(14, 26,12).

      3. The matrix ∇2C of second order partial derivatives of the cost function C(q;wL,wK,wM) = h(q)*c(wL,wK,wM) is symmetric.

      4. The cost function C(q;wL,wK,wM) is a concave in factor prices if the Hessian matrix H = ∇2wwc(wL,wK,wM) of second order partial derivatives with respect to factor prices is negative semidefinite.   The values of the second order partial derivatives depend on factor prices. As an example, consider the case where wL = 7, wK = 13, and wM = 6:

H = ∇2wwc =
-0.089420.025890.04823
0.02589-0.02090.01508
0.048230.01508-0.08895

The principal minors of H are H1 = -0.089425, H2 = 0.001199, and H3 = -0.

If these principal minors alternate in sign, starting with negative, with H3 <= 0, the matrix H is negative (semi)definite,
and the cost function C(q;wL,wK,wM) is a concave function in factor prices at wL = 7, wK = 13, and wM = 6.

The eigenvalues of H are e1 = -0.138, e2 = -0.0613, and e3 = 0.
H3 = e1 * e2 * e3 = 0.

XIV. Table of Results: check that the estimated Diewert cost function and input amounts for a given level of output agree with the Generalized CES production function's minimized cost and inputs. Also, compare the Allen partial elasticities of substitution, sLK, sLM, and sMK, with the Uzawa partial elasticities of substitution, uLK, uLM, and uKM.

Generalized CES-Diewert Cost Function
À2: Estimate the Diewert cost function directly
Parameters: rho = 0.17647, rhoL = 0.17647, rhoK = 0.11823, rhoM = 0.26339
 nu = 1
  —   Generalized CES Cost Data   —  
nu1 = 0.92
  —   Diewert Cost Data   —  
Factor SharesUzawa Elasticities2wwc(w)
obs #qwLwKwM LK MsLKsLMsKMεLKMcostest costest Lest Kest MsLsKsMuLKuLMuKLuKMuMLuMKe1e2e3
1188 13.027.02 24.2814.61 22.740.90.790.830.93544.08542.724.3914.7522.170.3590.3540.2870.850.970.850.540.970.54 -0.117-0.055-0
2197.46 11.287.92 26.3116.97 21.080.90.790.840.934554.62553.2326.7416.6720.920.3610.340.2990.820.990.820.550.990.55 -0.116-0.054-0
3208.86 11.84.44 22.5216.15 33.040.890.790.830.922536.81536.7822.0716.7632.310.3640.3680.2670.910.970.910.470.970.47 -0.171-0.060
4216.34 11.584.1 27.7815.18 32.830.890.790.830.92486.47487.5427.5915.8331.540.3590.3760.2650.880.940.880.520.940.52 -0.188-0.0760
5227.26 12.57.44 31.8118.46 25.870.90.790.830.929654.13653.4932.2118.3225.630.3580.350.2920.830.970.830.550.970.55 -0.122-0.0540
6236.98 12.44.86 30.8917.4 34.180.890.790.830.921597.44598.1630.7517.8833.30.3590.3710.2710.870.950.870.520.950.52 -0.159-0.068-0
7245.72 11.626.64 37.9219.15 27.60.90.790.830.927622.71622.1638.4619.0627.220.3540.3560.290.810.950.810.580.950.58 -0.154-0.06-0
8256.46 11.947.04 37.8920.82 290.90.790.830.927697.52697.4138.4320.5828.890.3560.3520.2920.820.960.820.560.960.56 -0.137-0.057-0
9268.02 12.567.92 36.6623.32 30.40.90.790.830.927827.61828.5137.1822.830.790.360.3460.2940.830.980.830.540.980.54 -0.111-0.0520
10276.9 14.94.84 39.2718.81 42.490.890.790.830.915756.89759.0939.119.541.050.3550.3830.2620.870.920.870.530.920.53 -0.165-0.068-0
11287.82 11.685.86 36.7624.4 37.920.890.790.830.922794.59795.2436.7724.2238.380.3620.3560.2830.860.970.860.520.970.52 -0.131-0.0610
12297.96 12.666.24 39.3524.78 38.990.890.790.830.921870.3871.0339.4224.6139.380.360.3580.2820.860.970.860.520.970.52 -0.126-0.0580
13307 136 43.7924.14 40.130.890.790.830.92861.17861.7643.9524.1140.120.3570.3640.2790.850.950.850.540.950.54 -0.138-0.0610
14318 14.024.52 40.7123.55 52.150.890.790.830.913891.55892.6340.1924.1951.330.360.380.260.90.940.90.50.940.5 -0.171-0.0620
15328.92 11.786.42 40.4830.15 42.830.890.790.830.922991.2992.7840.4829.4544.350.3640.3490.2870.870.990.870.50.990.5 -0.117-0.0570
16335 11.587.98 60.1527.67 32.970.90.790.830.926884.25884.6161.6926.8433.250.3490.3510.30.780.950.780.610.950.61 -0.178-0.051-0
17346.18 14.484.04 51.1823.11 57.490.890.790.830.909883.12885.2850.823.9255.680.3550.3910.2540.880.90.880.530.90.53 -0.199-0.0780
18357.36 12.627.42 52.3131.26 41.650.890.790.830.9221088.511090.5853.0430.3242.80.3580.3510.2910.830.970.830.550.970.55 -0.121-0.0540
19365.48 147.9 66.6628.24 39.320.890.790.830.9211071.321070.9867.7527.7139.470.3470.3620.2910.790.940.790.610.940.61 -0.165-0.0490
20375.02 12.587.38 68.5529.59 39.740.890.790.830.9211009.591009.4869.7828.8940.060.3470.360.2930.790.940.790.610.940.61 -0.179-0.053-0
21388.36 12.424.9 48.5432.84 59.660.890.790.830.9141106.031104.8747.8932.6960.910.3620.3670.270.890.960.890.490.960.49 -0.154-0.061-0
22395.4 14.424.36 65.0126.44 60.780.890.790.830.909997.32997.6864.7226.9959.410.350.390.260.860.90.860.560.90.56 -0.196-0.082-0
23408.02 14.384.58 54.3631.28 67.690.890.790.830.9091195.731194.1553.5731.6267.640.360.3810.2590.890.940.890.510.940.51 -0.169-0.0620
24415.3 13.944.14 67.8628.04 64.940.890.790.830.9081019.321019.0367.4728.5563.640.3510.3910.2590.860.90.860.560.90.56 -0.204-0.085-0
25425.84 11.584.14 62.0432.85 64.510.890.790.830.9111009.721008.2561.6132.8164.850.3570.3770.2660.870.930.870.530.930.53 -0.191-0.08-0
26437.64 13.744.84 60.1834.72 68.540.890.790.830.911268.541265.9859.5134.6469.290.3590.3760.2650.880.940.880.510.940.51 -0.159-0.063-0
27446.88 14.785.3 68.2533.81 66.010.890.790.830.911319.091316.6367.9133.6966.330.3550.3780.2670.860.930.860.540.930.54 -0.153-0.065-0
28457.16 12.988 71.6341.4 51.450.890.790.830.9191461.851465.0772.9239.553.780.3560.350.2940.820.970.820.560.970.56 -0.122-0.0510
29466.94 11.264.84 63.8740.52 67.140.890.790.830.9121224.541222.4763.4839.5969.470.360.3650.2750.870.960.870.510.960.51 -0.158-0.069-0
30475.86 14.784.1 7632.79 78.740.890.790.830.9051252.881249.9175.2333.2277.570.3530.3930.2540.870.90.870.550.90.55 -0.199-0.080
31487.6 12.46.06 68.9743.71 64.860.890.790.830.9141459.231458.3469.0342.1667.80.360.3590.2820.860.960.860.530.960.53 -0.13-0.06-0
AVE:336.96 12.865.85 49.1726.51 46.430.890.790.830.918933.49933.4149.1726.5146.430.3570.3660.2770.850.950.850.540.950.54-0.155-0.063-0



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).

À3:   Estimate the factor demand equations separately, with nu1 = 1.

XVI. For these coefficients of the CES Generalized production function, I generated a sequence (displayed in the "Generalized CES-Diewert Cost Function" table) of factor prices, outputs, and the corresponding cost minimizing inputs. Then I used these data to estimate the coefficients of each factor share equation separately:

QR Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
cLL-0.2605840.064-4.043
dLK0.8301590.04817.415
dLM0.6441990.04713.649
R2 = 0.9644 R2b = 0.9619 # obs = 31

QR Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
cKK0.0843180.0651.302
dKL0.5615380.0856.591
dKM0.4621830.0647.238
R2 = 0.83 R2b = 0.8179 # obs = 31

QR Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
cMM-0.2399830.008-28.546
dML0.6756670.01163.072
dMK0.5791620.00876.393
R2 = 0.9993 R2b = 0.9992 # obs = 31

The three estimated factor demand functions are:

∂C/∂wL = L(q; wL, wK, wM) = q^(1/1) * [-0.2606 + 0.8302 * (wK / wL)^(1/2) + 0.6442 * (wM / wL)^(1/2)]

∂C/∂wK = K(q; wL, wK, wM) = q^(1/1) * [0.0843 + 0.5615 * (wL / wK)^(1/2) + 0.4622 * (wM / wK)^(1/2)]

∂C/∂wM = M(q; wL, wK, wM) = q^(1/1) * [-0.24 + 0.6757 * (wL / wM)^(1/2) + 0.5792 * (wK / wM)^(1/2)]

As estimated, generally dLK != dKL, dLM != dML, and dKM != dMK: Young's Theorem doesn't hold without constraints across equations. Consequently, I use the sum of the cross parameters in the Diewert cost function.

    The estimated factor demand elasticities are obtained by:

εL,wL = ∂ln(L(q;wL,wK,wM))/∂ln(wL) = -.5 * (q^1/nu1) / L(q;wL,wK,wM)) * (dLK * (wK/wL)^1/2 + dLM * (wM/wL)^1/2),

εL,wK = ∂ln(L(q;wL,wK,wM))/∂ln(wK) = .5 * (q^1/nu1) / L(q;wL,wK,wM)) * dLK * (wK/wL)^1/2,

εL,wM = ∂ln(L(q;wL,wK,wM))/∂ln(wM) = .5 * (q^1/nu1) / L(q;wL,wK,wM)) * dLM * (wM/wL)^1/2,

εL,q = ∂ln(L(q;wL,wK,wM))/∂ln(q) = 1/nu1, etc.

      For example, with wL = 7, wK = 13, wM = 6, and q = 30:

εL,wLεL,wKεL,wMεL,q
εK,wLεK,wKεK,wMεK,q
εM,wLεM,wKεM,wMεM,q
-0.58880.38560.20331
0.2542-0.4480.19371
0.27180.3175-0.58941

XVII. The Diewert cost function as obtained from the unrestricted factor demand equations is:

C(q;wL,wK,wM) = q^(1/nu1) * [cLL * wL + cKK * wK + cMM * wM + (dLK+dKL) * (wL*wK)^(1/2) + (dLM +dML) * (wL*wM)^(1/2) + (dKM+dMK) * (wK*wM)^(1/2)]

= q^(1/1) * [-0.26058 * wL + 0.08432 * wK + -0.23998 * wM + 1.3917 * (wL*wK)^(1/2) + 1.31987 * (wL*wM)^(1/2) + 1.04134 * (wK*wM)^(1/2)]
        (***)

XVIII. The Restricted Factor Demand Equations.

Having obtained the unrestricted estimates of the coefficients of the factor demand functions, we compute the restricted least squares estimates with dLK = dKL, dLM = dML, and dKM = dMK.

QR Restricted Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
cLL-0.1983390.045355-4.373002
dLK0.7515950.03154323.827923
dLM0.6937820.02730125.412061
cKK-0.1115280.029193-3.820419
dKL0.7515950.03154323.827923
dKM0.5454640.0232923.420232
cMM-0.2089790.033151-6.303848
dML0.6937820.02730125.412061
dMK0.5454640.0232923.420232
R2 = 0.9945 R2b = 0.994 # obs = 93

dLK = dKL, dLM = dML, and dKM = dMK

The three estimated, restricted factor demand functions are:

∂C/∂wL = L(q; wL, wK, wM) = q^(1/1) * [-0.1983 + 0.7516 * (wK / wL)^(1/2) + 0.6938 * (wM / wL)^(1/2)]

∂C/∂wK = K(q; wL, wK, wM) = q^(1/1) * [-0.1115 + 0.7516 * (wL / wK)^(1/2) + 0.5455 * (wM / wK)^(1/2)]

∂C/∂wM = M(q; wL, wK, wM) = q^(1/1) * [-0.209 + 0.6938 * (wL / wM)^(1/2) + 0.5455 * (wK / wM)^(1/2)]

    The estimated factor demand elasticities are obtained by:

εL,wL = ∂ln(L(q;wL,wK,wM))/∂ln(wL) = -.5 * (q^1/nu1) / L(q;wL,wK,wM)) * (dLK * (wK/wL)^1/2 + dLM * (wM/wL)^1/2),

εL,wK = ∂ln(L(q;wL,wK,wM))/∂ln(wK) = .5 * (q^1/nu1) / L(q;wL,wK,wM)) * dLK * (wK/wL)^1/2,

εL,wM = ∂ln(L(q;wL,wK,wM))/∂ln(wM) = .5 * (q^1/nu1) / L(q;wL,wK,wM)) * dLM * (wM/wL)^1/2,

εL,wq = ∂ln(L(q;wL,wK,wM))/∂ln(q) = 1/nu, etc.

      For example, with wL = 7, wK = 13, wM = 6, and q = 30:

εL,wLεL,wKεL,wMεL,q
εK,wLεK,wKεK,wMεK,q
εM,wLεM,wKεM,wMεM,q
-0.56750.34880.21871
0.3402-0.56880.22861
0.27890.2989-0.57781

XIX. The Diewert cost function as obtained from the restricted factor demand equations is:

C(q;wL,wK,wM) = q^(1/nu) * [cLL * wL + cKK * wK + cMM * wM + (dLK+dKL) * (wL*wK)^(1/2) + (dLM +dML) * (wL*wM)^(1/2) + (dKM+dMK) * (wK*wM)^(1/2)]

= q^(1/1) * [-0.19834 * wL + -0.11153 * wK + -0.20898 * wM + 1.50319 * (wL*wK)^(1/2) + 1.38756 * (wL*wM)^(1/2) + 1.09093 * (wK*wM)^(1/2)]
        (***)

XX. Note:
      1. C(q;wL,wK,wM) linear homogeneous in factor prices requires for any t > 0 that the unit cost function c(wL,wK,wM) obey:

c(t*wL,t*wK,t*wM) = t * c(wL,wK,wM).

        For example, with wL = 7, wK = 13, wM = 6, and t = 2:

2 * c(7, 13, 6) = 57.75 = 57.75 = c(14, 26,12).

      2. The matrix ∇2C of second order partial derivatives of the cost function C(q;wL,wK,wM) = h(q)*c(wL,wK,wM) is symmetric:

2C/∂q∂wL = h'(q) * ∂c/∂wL = ∂2C/∂wL∂q,

2C/∂q∂wK = h'(q) * ∂c/∂wK = ∂2C/∂wK∂q,

2C/∂q∂wM = h'(q) * ∂c/∂wM = ∂2C/∂wM∂q,

2c/∂wL∂wK = .25 * (dLK+dKL)/(wL*wK)^(1/2) = ∂2c/∂wK∂wL,

2c/∂wL∂wM = .25 * (dLM+dML)/(wL*wM)^(1/2) = ∂2c/∂wM∂wL,

2c/∂wK∂wM = .25 * (dKM+dMK)/(wL*wK)^(1/2) = ∂2c/∂wM∂wK, and

2c/∂wL∂wL = -.25((dLK+dKL)*wK^(1/2) + (dLM+dML)*wM^(1/2))/wL^(3/2),

2c/∂wK∂wK = -.25((dKL+dLK)*wL^(1/2) + (dKM+dMK)*wM^(1/2))/wK^(3/2),

2c/∂wM∂wM = -.25((dML+dLM)*wL^(1/2) + (dMK+dKM)*wK^(1/2))/wM^(3/2).

      3. The cost function C(q;wL,wK,wM) is a concave in factor prices if the Hessian matrix H = ∇2wwc(wL,wK,wM) of second order partial derivatives with respect to factor prices is negative semidefinite.   The values of the second order partial derivatives depend on factor prices. As an example, consider the case where wL = 7, wK = 13, and wM = 6:

H = ∇2wwc =
-0.119040.039390.05353
0.03939-0.035470.03088
0.053530.03088-0.12936

The principal minors of H are H1 = -0.119041, H2 = 0.00267, and H3 = -0.

If these principal minors alternate in sign, starting with negative, with H3 <= 0, the matrix H is negative (semi)definite,
and the cost function C(q;wL,wK,wM) is a concave function in factor prices at wL = 7, wK = 13, and wM = 6.

The eigenvalues of H are e1 = -0.1781, e2 = -0.1058, and e3 = -0.
H3 = e1 * e2 * e3 = -0.

XXI. The factor share functions are:

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

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

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

XXII. Uzawa Partial Elasticities of Substitution:

uLK = C(q;wL,wK,wM) * ∂L(q;wL,wK,wM)/∂wK / (L(q;wL,wK,wM) * K(q;wL,wK,wM)),

uLM = C(q;wL,wK,wM) * ∂L(q;wL,wK,wM)/∂wM / (L(q;wL,wK,wM) * M(q;wL,wK,wM)),

uKL = C(q;wL,wK,wM) * ∂K(q;wL,wK,wM)/∂wL / (K(q;wL,wK,wM) * L(q;wL,wK,wM)),

uKM = C(q;wL,wK,wM) * ∂K(q;wL,wK,wM)/∂wM / (K(q;wL,wK,wM) * M(q;wL,wK,wM)),

uML = C(q;wL,wK,wM) * ∂M(q;wL,wK,wM)/∂wL / (M(q;wL,wK,wM) * L(q;wL,wK,wM)),

uMK = C(q;wL,wK,wM) * ∂M(q;wL,wK,wM)/∂wK / (M(q;wL,wK,wM) * M(q;wL,wK,wM)),

Imposing the dLK = dKL, dLM = dML, and dKM = dMK restrictions in the factor demand functions, we get (approximate?) equality of the cross partial elasticities of substitution:

uLK = uKL,    uLM = uML,  and    uKM = uMK.

as seen in the following table.

XXIII. Table of Results: check that the estimated Diewert cost function and input amounts for a given level of output agree with the Generalized CES production function's minimized cost and inputs. Also, compare the Allen partial elasticities of substitution, sLK, sLM, and sMK, with the Uzawa partial elasticities of substitution, uLK, uLM, and uKM.

Generalized CES-Diewert Cost Function
À3: Estimate the Translog cost function using restricted factor shares
Parameters: rho = 0.17647, rhoL = 0.17647, rhoK = 0.11823, rhoM = 0.26339
 nu = 1
  —   Generalized CES Cost Data   —  
nu1 = 1
  —   Diewert Cost Data   —  
Factor SharesUzawa Elasticities2wwc(w)
obs #qwLwKwM LK MsLKsLMsKMεLKMcostest costest Lest Kest MsLsKsMuLKuLMuKLuKMuMLuMKe1e2e3
1188 13.027.02 24.2814.61 22.740.90.790.830.93544.08569.9425.3915.8122.940.3560.3610.2830.940.820.940.810.820.81 -0.15-0.095-0
2197.46 11.287.92 26.3116.97 21.080.90.790.840.934554.62577.0927.3718.1821.190.3540.3550.2910.90.850.90.820.850.82 -0.147-0.0940
3208.86 11.84.44 22.5216.15 33.040.890.790.830.922536.81559.3623.217.4933.210.3680.3690.2641.010.81.010.730.80.73 -0.233-0.0960
4216.34 11.584.1 27.7815.18 32.830.890.790.830.92486.47505.3828.8816.1532.980.3620.370.26810.7610.790.760.79 -0.254-0.1250
5227.26 12.57.44 31.8118.46 25.870.90.790.830.929654.13674.2932.7819.4126.030.3530.360.2870.920.820.920.830.820.83 -0.155-0.094-0
6236.98 12.44.86 30.8917.4 34.180.890.790.830.921597.44615.331.7918.2634.360.3610.3680.2710.980.770.980.790.770.79 -0.212-0.1130
7245.72 11.626.64 37.9219.15 27.60.90.790.830.927622.71637.738.8919.8827.760.3490.3620.2890.910.80.910.860.80.86 -0.198-0.105-0
8256.46 11.947.04 37.8920.82 290.90.790.830.927697.52711.9338.6921.529.150.3510.3610.2880.920.810.920.840.810.84 -0.174-0.0990
9268.02 12.567.92 36.6623.32 30.40.90.790.830.927827.61841.8737.2223.9830.580.3550.3580.2880.920.840.920.820.840.82 -0.14-0.090
10276.9 14.94.84 39.2718.81 42.490.890.790.830.915756.89769.0440.1519.1942.560.360.3720.26810.7310.820.730.82 -0.221-0.113-0
11287.82 11.685.86 36.7624.4 37.920.890.790.830.922794.59803.7736.9824.9138.150.360.3620.2780.960.820.960.780.820.78 -0.17-0.1040
12297.96 12.666.24 39.3524.78 38.990.890.790.830.921870.3877.8539.5525.1539.20.3590.3630.2790.960.810.960.790.810.79 -0.163-0.0990
13307 136 43.7924.14 40.130.890.790.830.92861.17866.2444.0524.3240.30.3560.3650.2790.960.780.960.820.780.82 -0.178-0.106-0
14318 14.024.52 40.7123.55 52.150.890.790.830.913891.55894.4440.8623.7451.910.3650.3720.2621.010.751.010.770.750.77 -0.234-0.10
15328.92 11.786.42 40.4830.15 42.830.890.790.830.922991.2991.0940.1330.2543.130.3610.3590.2790.960.840.960.760.840.76 -0.152-0.0960
16335 11.587.98 60.1527.67 32.970.90.790.830.926884.25882.3960.1227.5632.910.3410.3620.2980.870.810.870.910.810.91 -0.23-0.089-0
17346.18 14.484.04 51.1823.11 57.490.890.790.830.909883.12877.6151.4422.757.180.3620.3750.2631.020.71.020.820.70.82 -0.271-0.128-0
18357.36 12.627.42 52.3131.26 41.650.890.790.830.9221088.511080.851.8930.8241.770.3530.360.2870.920.820.920.830.820.83 -0.154-0.0940
19365.48 147.9 66.6628.24 39.320.890.790.830.9211071.321060.9166.127.6639.420.3410.3650.2940.90.770.90.910.770.91 -0.213-0.0870
20375.02 12.587.38 68.5529.59 39.740.890.790.830.9211009.59997.5867.8128.939.790.3410.3640.2940.890.780.890.910.780.91 -0.232-0.0930
21388.36 12.424.9 48.5432.84 59.660.890.790.830.9141106.031088.3747.4632.2159.490.3650.3680.26810.7910.750.790.75 -0.208-0.0990
22395.4 14.424.36 65.0126.44 60.780.890.790.830.909997.32977.2264.4825.2960.650.3560.3730.27110.710.850.70.85 -0.26-0.140
23408.02 14.384.58 54.3631.28 67.690.890.790.830.9091195.731170.1753.2930.367.020.3650.3720.2621.010.751.010.770.750.77 -0.232-0.10
24415.3 13.944.14 67.8628.04 64.940.890.790.830.9081019.32993.7166.9826.6264.650.3570.3730.26910.710.850.70.85 -0.271-0.1440
25425.84 11.584.14 62.0432.85 64.510.890.790.830.9111009.72983.7660.6531.4364.150.360.370.270.990.750.990.810.750.81 -0.254-0.133-0
26437.64 13.744.84 60.1834.72 68.540.890.790.830.911268.541233.0758.5633.2268.010.3630.370.26710.7610.780.760.78 -0.215-0.1040
27446.88 14.785.3 68.2533.81 66.010.890.790.830.911319.091279.1666.5432.0365.660.3580.370.2720.980.740.980.820.740.82 -0.203-0.111-0
28457.16 12.988 71.6341.4 51.450.890.790.830.9191461.851420.6569.6139.3751.40.3510.360.2890.910.820.910.850.820.85 -0.156-0.089-0
29466.94 11.264.84 63.8740.52 67.140.890.790.830.9121224.541184.0361.5738.4666.870.3610.3660.2730.980.790.980.780.790.78 -0.209-0.116-0
30475.86 14.784.1 7632.79 78.740.890.790.830.9051252.881203.9374.0530.577.840.360.3740.2651.010.691.010.830.690.83 -0.27-0.1340
31487.6 12.46.06 68.9743.71 64.860.890.790.830.9141459.231406.8466.341.1964.720.3580.3630.2790.960.80.960.80.80.8 -0.169-0.103-0
AVE:336.96 12.865.85 48.826.34 46.290.890.790.830.918933.49926.9548.826.3446.290.3570.3660.2770.960.780.960.820.780.82-0.204-0.1060




À4:   Estimate the Diewert cost function directly, with nu1 = 1.

XXIV. If we have a data set relating the input (factor) prices, wL, wK, and wM, to the total (minimum) cost of producing output for varying levels output, but do not have data on the required levels of inputs, we can estimate the Diewert cost function directly. We will use the same sequence (displayed in the "Diewert Cost Function" table) of factor prices, outputs, and total (minimum) cost as used in À1. The estimated coefficients of the cost function will vary with the parameters nu, rho, rhoL, rhoK, rhoM, alpha, beta and gamma of the Generalized CES production function used to generate these data.

Under these circumstances, with nu1 = 1, the Diewert cost function may fail to be concave in factor prices, depending on the generated sequence of random factor prices, as one or more of the eigenvalues of the Hessian matrix H = ∇2wwc(wL,wK,wM) turn positive.

SVD Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
cLL-0.1201161.875-0.064
cKK-0.2943251.116-0.264
cMM-1.400521.808-0.775
2*dLK1.09232.7910.391
2*dLM1.3840471.7370.797
2*dKM2.567442.1691.184
R2 = 0.9238 R2b = 0.9085 # obs = 31
Observation Matrix Rank: 6
The Diewert cost function as obtained by direct estimation is:

C(q;wL,wK,wM) = q^(1/nu1) * [cLL * wL + cKK * wK + cMM * wM + 2 * dLK * (wL*wK)^(1/2) + 2 * dLM * (wL*wM)^(1/2) + 2 * dKM * (wK*wM)^(1/2)]

= q^(1/1) * [-0.12012 * wL + -0.29432 * wK + -1.40052 * wM + 1.0923 * (wL*wK)^(1/2) + 1.38405 * (wL*wM)^(1/2) + 2.56744 * (wK*wM)^(1/2)]
        (***)

XXVI. Its three derived factor demand functions are:

∂C/∂wL = L(q; wL, wK, wM) = q^(1/1) * [-0.1201 + 0.5461 * (wK / wL)^(1/2) + 0.692 * (wM / wL)^(1/2)]

∂C/∂wK = K(q; wL, wK, wM) = q^(1/1) * [-0.2943 + 0.5461 * (wL / wK)^(1/2) + 1.2837 * (wM / wK)^(1/2)]

∂C/∂wM = M(q; wL, wK, wM) = q^(1/1) * [-1.4005 + 0.692 * (wL / wM)^(1/2) + 1.2837 * (wK / wM)^(1/2)]

    The derived factor demand elasticities are obtained by:

εL,wL = ∂ln(L(q;wL,wK,wM))/∂ln(wL) = -.5 * (q^1/nu1) / L(q;wL,wK,wM)) * (dLK * (wK/wL)^1/2 + dLM * (wM/wL)^1/2),

εL,wK = ∂ln(L(q;wL,wK,wM))/∂ln(wK) = .5 * (q^1/nu1) / L(q;wL,wK,wM)) * dLK * (wK/wL)^1/2,

εL,wM = ∂ln(L(q;wL,wK,wM))/∂ln(wM) = .5 * (q^1/nu1) / L(q;wL,wK,wM)) * dLM * (wM/wL)^1/2,

εL,q = ∂ln(L(q;wL,wK,wM))/∂ln(q) = 1/nu, etc.

      For example, with wL = 7, wK = 13, wM = 6, and q = 30:

εL,wLεL,wKεL,wMεL,q
εK,wLεK,wKεK,wMεK,q
εM,wLεM,wKεM,wMεM,q
-0.54750.29420.25331
0.2048-0.65040.44561
0.30220.7641-1.06631

XXVII. Notes:
      1. As derived, dLK = dKL, dLM = dML, and dKM = dMK: Young's Theorem holds by construction.

      2. C(q;wL,wK,wM) linear homogeneous in factor prices requires for any t > 0 that the unit cost function c(wL,wK,wM) obey:

c(t*wL,t*wK,t*wM) = t * c(wL,wK,wM).

        For example, with wL = 7, wK = 13, wM = 6, and t = 2:

2 * c(7, 13, 6) = 57.99 = 57.99 = c(14, 26,12).

      3. The matrix ∇2C of second order partial derivatives of the cost function C(q;wL,wK,wM) = h(q)*c(wL,wK,wM) is symmetric.

      4. The cost function C(q;wL,wK,wM) is a concave in factor prices if the Hessian matrix H = ∇2wwc(wL,wK,wM) of second order partial derivatives with respect to factor prices is negative semidefinite.   The values of the second order partial derivatives depend on factor prices. As an example, consider the case where wL = 7, wK = 13, and wM = 6:

H = ∇2wwc =
-0.098930.028630.05339
0.02863-0.048960.07268
0.053390.07268-0.21975

The principal minors of H are H1 = -0.098926, H2 = 0.004024, and H3 = 0.

If these principal minors alternate in sign, starting with negative, with H3 <= 0, the matrix H is negative (semi)definite,
and the cost function C(q;wL,wK,wM) is a concave function in factor prices at wL = 7, wK = 13, and wM = 6.

The eigenvalues of H are e1 = -0.2573, e2 = -0.1103, and e3 = -0.
H3 = e1 * e2 * e3 = -0.

XXVIII. Table of Results: check that the estimated Diewert cost function and input amounts for a given level of output agree with the Generalized CES production function's minimized cost and inputs. Also, compare the Allen partial elasticities of substitution, sLK, sLM, and sMK, with the Uzawa partial elasticities of substitution, uLK, uLM, and uKM.

Generalized CES-Diewert Cost Function
À4: Estimate the Diewert cost function directly
Parameters: rho = 0.17647, rhoL = 0.17647, rhoK = 0.11823, rhoM = 0.26339
 nu = 1
  —   Generalized CES Cost Data   —  
nu1 = 1
  —   Diewert Cost Data   —  
Factor SharesUzawa Elasticities2wwc(w)
obs #qwLwKwM LK MsLKsLMsKMεLKMcostest costest Lest Kest MsLsKsMuLKuLMuKLuKMuMLuMKe1e2e3
1188 13.027.02 24.2814.61 22.740.90.790.830.93544.08565.9422.0519.3819.560.3120.4460.2430.641.090.641.81.091.8 -0.215-0.099-0
2197.46 11.287.92 26.3116.97 21.080.90.790.840.934554.62562.7324.0323.2815.260.3190.4670.2150.571.310.572.041.312.04 -0.187-0.112-0
3208.86 11.84.44 22.5216.15 33.040.890.790.830.922536.81553.552019.3333.40.320.4120.2680.760.910.761.520.911.52 -0.377-0.0850
4216.34 11.584.1 27.7815.18 32.830.890.790.830.92486.47508.0924.6618.3533.970.3080.4180.2740.750.860.751.590.861.59 -0.404-0.1160
5227.26 12.57.44 31.8118.46 25.870.90.790.830.929654.13668.0528.5424.4720.830.310.4580.2320.61.160.61.921.161.92 -0.2-0.111-0
6236.98 12.44.86 30.8917.4 34.180.890.790.830.921597.44617.7727.2621.1434.020.3080.4240.2680.720.910.721.630.911.63 -0.331-0.107-0
7245.72 11.626.64 37.9219.15 27.60.90.790.830.927622.71637.9333.6925.4222.560.3020.4630.2350.61.130.61.951.131.95 -0.228-0.140
8256.46 11.947.04 37.8920.82 290.90.790.830.927697.52707.933.6227.3323.350.3070.4610.2320.61.160.61.941.161.94 -0.212-0.125-0
9268.02 12.567.92 36.6623.32 30.40.90.790.830.927827.61828.0632.5330.223.720.3150.4580.2270.61.210.61.931.211.93 -0.187-0.102-0
10276.9 14.94.84 39.2718.81 42.490.890.790.830.915756.89779.8734.0821.8445.310.3010.4170.2810.760.820.761.610.821.61 -0.348-0.107-0
11287.82 11.685.86 36.7624.4 37.920.890.790.830.922794.59797.0232.129.7333.920.3150.4360.2490.671.050.671.721.051.72 -0.262-0.1-0
12297.96 12.666.24 39.3524.78 38.990.890.790.830.921870.3873.4134.2630.1635.080.3120.4370.2510.671.030.671.731.031.73 -0.246-0.0980
13307 136 43.7924.14 40.130.890.790.830.92861.17869.8337.9529.3637.10.3050.4390.2560.670.990.671.740.991.74 -0.257-0.11-0
14318 14.024.52 40.7123.55 52.150.890.790.830.913891.55896.2534.8126.2655.210.3110.4110.2780.780.830.781.550.831.55 -0.38-0.09-0
15328.92 11.786.42 40.4830.15 42.830.890.790.830.922991.2974.9835.0336.1236.930.320.4360.2430.661.10.661.731.11.73 -0.237-0.09-0
16335 11.587.98 60.1527.67 32.970.90.790.830.926884.25876.1452.3137.322.890.2990.4930.2080.531.320.532.261.322.26 -0.21-0.153-0
17346.18 14.484.04 51.1823.11 57.490.890.790.830.909883.12891.5943.3625.1864.110.3010.4090.2910.80.760.81.580.761.58 -0.437-0.119-0
18357.36 12.627.42 52.3131.26 41.650.890.790.830.9221088.511071.3345.1538.7533.70.310.4560.2330.611.150.611.91.151.9 -0.201-0.11-0
19365.48 147.9 66.6628.24 39.320.890.790.830.9211071.321073.9557.0136.4231.850.2910.4750.2340.581.120.582.031.122.03 -0.206-0.141-0
20375.02 12.587.38 68.5529.59 39.740.890.790.830.9211009.591006.4558.5938.2531.310.2920.4780.230.571.150.572.071.152.07 -0.221-0.1540
21388.36 12.424.9 48.5432.84 59.660.890.790.830.9141106.031082.8240.8636.4858.790.3150.4180.2660.740.930.741.580.931.58 -0.333-0.09-0
22395.4 14.424.36 65.0126.44 60.780.890.790.830.909997.321002.8154.3729.0966.460.2930.4180.2890.770.770.771.640.771.64 -0.391-0.141-0
23408.02 14.384.58 54.3631.28 67.690.890.790.830.9091195.731173.7745.3733.5271.60.310.4110.2790.790.830.791.550.831.55 -0.376-0.090
24415.3 13.944.14 67.8628.04 64.940.890.790.830.9081019.321018.3956.4730.4271.260.2940.4160.290.770.770.771.630.771.63 -0.414-0.1430
25425.84 11.584.14 62.0432.85 64.510.890.790.830.9111009.72993.5951.7336.1765.870.3040.4220.2740.740.860.741.620.861.62 -0.397-0.127-0
26437.64 13.744.84 60.1834.72 68.540.890.790.830.911268.541238.5950.0137.6270.170.3080.4170.2740.750.860.751.590.861.59 -0.344-0.096-0
27446.88 14.785.3 68.2533.81 66.010.890.790.830.911319.091297.8556.6637.2767.390.30.4240.2750.730.860.731.650.861.65 -0.308-0.1090
28457.16 12.988 71.6341.4 51.450.890.790.830.9191461.851407.7560.650.3640.020.3080.4640.2270.591.190.591.981.191.98 -0.186-0.1140
29466.94 11.264.84 63.8740.52 67.140.890.790.830.9121224.541182.4253.0644.963.760.3110.4280.2610.710.960.711.650.961.65 -0.327-0.109-0
30475.86 14.784.1 7632.79 78.740.890.790.830.9051252.881228.5662.3334.1187.610.2970.410.2920.80.750.81.590.751.59 -0.429-0.127-0
31487.6 12.46.06 68.9743.71 64.860.890.790.830.9141459.231401.7457.3849.4758.120.3110.4380.2510.671.030.671.731.031.73 -0.253-0.102-0
AVE:336.96 12.865.85 41.9331.54 44.360.890.790.830.918933.49928.6841.9331.5444.360.3070.4370.2560.6810.681.7611.76-0.294-0.114-0



Mathematical Notes

1. The Diewert (Generalized Leontief) cost function

C(q;wL,wK,wM) = h(q) * c(wL,wK,wM)         (**)

where the returns to scale function is: h(q) = q^(1/nu), a continuous, increasing function of q (q >= 1), with h(0) = 0 and h(1) = 1,

and the unit cost function is: c(wL,wK,wM) = cLL * wL + cKK * wK + cMM * wM + (dLK+dKL) * (wL*wK)^(1/2) + (dLM +dML) * (wL*wM)^(1/2) + (dKM+dMK) * (wK*wM)^(1/2),

linear in its parameters cLL, cKK, cMM, dLK, dKL, dLM, dML, dKM, and dMK.

2. The Factor Demand Functions:

∂C/∂wL = L(q; wL, wK, wM) = q^(1/nu) * [cLL + dLK * (wK / wL)^(1/2) + dLM * (wM / wL)^(1/2)]

∂C/∂wK = K(q; wL, wK, wM) = q^(1/nu) * [cKK + dKL * (wL / wK)^(1/2) + dKM * (wM / wK)^(1/2)]

∂C/∂wM = M(q; wL, wK, wM) = q^(1/nu) * [cMM + dML * (wL / wM)^(1/2) + dMK * (wK / wM)^(1/2)]

3. The Factor Share Functions:

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

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

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

4. The Factor Demand Elasticities:

εL,wL = ∂ln(L(q;wL,wK,wM))/∂ln(wL) = -.5 * (q^1/nu) / L(q;wL,wK,wM)) * (dLK * (wK/wL)^1/2 + dLM * (wM/wL)^1/2),

εL,wK = ∂ln(L(q;wL,wK,wM))/∂ln(wK) = .5 * (q^1/nu) / L(q;wL,wK,wM)) * dLK * (wK/wL)^1/2,

εL,wM = ∂ln(L(q;wL,wK,wM))/∂ln(wM) = .5 * (q^1/nu) / L(q;wL,wK,wM)) * dLM * (wM/wL)^1/2,

εL,q = ∂ln(L(q;wL,wK,wM))/∂ln(q) = 1/nu, etc.

5. Cost Function C(q;wL,wK,wM) Concave in Factor Prices:

2c/∂wL∂wL = -.25((dLK+dKL)*wK^(1/2) + (dLM+dML)*wM^(1/2))/wL^(3/2),

2c/∂wL∂wK = .25 * (dLK+dKL)/(wL*wK)^(1/2) = ∂2c/∂wK∂wL,

2c/∂wL∂wM = .25 * (dLM+dML)/(wL*wM)^(1/2) = ∂2c/∂wM∂wL,

2c/∂wK∂wK = -.25((dKL+dLK)*wL^(1/2) + (dKM+dMK)*wM^(1/2))/wK^(3/2),

2c/∂wK∂wM = .25 * (dKM+dMK)/(wL*wK)^(1/2) = ∂2c/∂wM∂wK, and

2c/∂wM∂wM = -.25((dML+dLM)*wL^(1/2) + (dMK+dKM)*wK^(1/2))/wM^(3/2).

 

 
   

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