International Journal of Economics and Financial

Issues
ISSN: 2146-4138

available at http: www.econjournals.com

International Journal of Economics and Financial Issues, 2018, 8(3), 142-147.

Testing the Fit of Cobb-Douglass Production Function within

Unrestricted Least Squares

Joseph I. Amuka1*, Fredrick O. Asogwa2, Collins C. Ugwu3, Kelvin C. Ugwu4

Department of Economics, University of Nigeria Nsukka, Enugu, Nigeria, 2Department of Economics, University of Nigeria
1

Nsukka, Enugu, Nigeria, 3Department of Accountancy, Federal University Wukara, Taraba, Nigeria, 4Department of Economics,
University of Nigeria Nsukka, Enugu, Nigeria. *Email: [email protected]

ABSTRACT
Criticisms trailed Cobb-Douglass production model and its constant returns to Scale assumption after its application in America Economy in 1927.
Studies using restricted least squares approach to validate the model have produced different results in different economies. The question this paper
tried to answer is this-without restriction, can Cobb-Douglass model still fit well in an economy’s production? We answered this question using
unrestricted least squares method with quarterly data generated from Central Bank of Nigeria between 2009 and 2012. The study is very important
in Nigeria where the economy has not performed well despite heavy government investment in the last two decades. Result from the shows that even
without restriction, Nigeria economy displays constant returns to scale production, suggesting that the Cobb-Douglass Production Function fits well
in Nigeria economy. The result has an implication for a country in which the economy is driven by the public sector.
Keywords: Cobb-Douglass, Production Function, Constant Returns, Unrestricted Least Square
JEL Classification: E23

1. INTRODUCTION in the real sense of it. The neoclassical economists also attacked
the model on the bases that the productivity theory is more of
The Cobb-Douglass theory of production has provided important an abstraction than a quantifiable. Douglass while responding
framework for the measurement of productivity and employment to the criticisms noted that the critics were so hostile even to the
of factors of production since 1930s. Cobb and Douglas have extent of recommending for the work to be thrown into the waste
modelled the growth of output in American manufacturing sector basket and further research in it stopped. In the defence of the
between 1899 and 1922 in which output of goods were determined model, From criticism, Cobb-Douglass model started receiving
by combination of two factor inputs, namely labour and capital research interest with positive comments and positive empirical
under the assumption of constant returns to scale production. result. Miller (2008) accepts that Cobb-Douglass model is very
Research attention given to the novel work of the Cobb-Douglass simple to use and can fit many data sets very well for empirical
model in the last five decades is a worthwhile because of its forecasting. Many studies have equally been done in developing as
significance to the macro and micro economy. Following the well as developed countries, trying to validate the Cobb-Douglass
assumptions made in the model, the work of Cobb and Douglass model. Results of such studies have differed, making it difficult to
attracted many criticisms and questions. make a definite conclusion about the Cobb-Douglass postulation.
Hence, up to this point, the applicability of constant returns to scale
Fraser (2002) saw the omission of technical change in the production as postulated by Cobb-Douglass is still at the centre
specification of the Cobb-Douglass model as a serious limitation to of research interest around the globe.
the acceptance of its assumptions. By failure to recognize technical
change, Fraser says Cobb and Douglass assumed that technology Adetunji et al. (2012); Abidemi (2010) and some other studies have
was constant within the period of their study, which does not hold been done to validate the applicability of Cobb-Douglass laws of

142 International Journal of Economics and Financial Issues | Vol 8 • Issue 3 • 2018
 Amuka, et al.: Testing the Fit of Cobb-Douglass Production Function within Unrestricted Least Squares

production with Nigerian data. Adetunji et al. used macro data made the first functional relationship between production inputs
to study the application of laws of production as propounded by and outputs, but could not apply it before the novel work of
cobb-Douglass model. Their result shows that production function Charles Cobb and Paul Douglas. Cobb and Douglass came out
in Nigeria follows the constant returns to scale as predicted by with the law of production after conducting an empirical research
cobb-Douglass model, which means that doubling input use in America manufacturing sector to show that the production of
will double output in the country. However, in a micro study by a given quantity of good depends on the combination of some
Abidemi (2010), result shows that in the banking sector, when factor inputs, capital and labour. They became the first group of
inputs are doubled, output will more than double in the sector. It scholars to model aggregate production function and quantify
then suggests that in the banking industry in Nigeria increasing the marginal product of any factor in economic literature. From
returns to scale prevails. This contradicting result between micro then, there emerged production function which defines a technical
and macro studies is a source of worry. relationship between output and inputs in production process. As
fundamentally expressed by cobb-Douglass,
The objective of this study is to remove the restriction imposed by
the Cobb-Douglas model and find out if it can still fit the economy Y = ALαKβ (1)
of Nigeria. The removal of the restriction is one of the steps taken
to reconcile the macro and micro studies done in Nigeria so as to Where Y = total output, L = units of labour, K = units of capital, and
provide policy makers a more reliable insight of the application α and β are elasticity of labour and capital, and A is an efficiency
of the model in the country. In order to make the reconciliation, parameter. Cobb-Douglass stated that the production function
we decided to use study methodology which is different from follows constant returns to scale, which denotes that when factor
the earlier ones. Previous studies used restricted least squares as inputs are increased by 1%, output will equally increase by 1%
empirical model. We applied Unrestricted least squares in a macro (α+β = 1).
data and make suggestion that another study should apply the
same unrestricted least squares in a micro data. Moreover, rather Cobb-Douglas model received two major criticisms from scholars
than use government investment as a proxy of total investment, of economics based on its assumptions. One of the criticisms
we used gross fixed capital formation (GFCF) to represent the centred on assumption of constant returns to scale which the model
economy’s total investment. built its analysis on. The second one and the more important is the
omission of technical change, thereby believing that technology
In order to achieve our objective, the remainder of the work is remained the same within the period of the study. Fraser (2002)
organised as follows: Section 11 is the theoretical framework and regarded the restriction imposed in the model without relevant tests
literature review; Section 111 describes the methodology used; as econometrically unacceptable and their belief that technology
Section 1V is the result and discussion of research findings and was constant as very difficult to accept. However, despite the
Section V is the conclusion the work. criticisms of the Cobb-Douglass production function, it is still
important in the theory of production today.
2. THEORETICAL FRAMEWORK AND The cobb-Douglass production function is very important to the
LITERATURE REVIEW business firm because it helps the firm to make rational decision
on the quantity of each factor inputs to employ so as to minimize
2.1. Theoretical Framework the production cost. The firm as a rational economic agent needs
Economic scholars in the 19th century did not pay much attention information on the marginal productivity of factors to be able to
to the mathematical modelling of the relationship between input produce at optimum. In trying to maximize profit or minimize loss,
and output. The physiocrats only recognised labour as a factor firms can substitute one factor for another under a cost outlay. For
of production and treated capital as non-productive because they instance if production is capital intensive, but the cost of capital is
never believed in value added by the manufacturing sector. The rising, decision can be taken to adjust inputs combination if rental
Physiocrats could not reason that the price difference between price of labour is unchanged. More labour can be combined with
processed and unprocessed goods was made possible because of less capital as long as they are the only factors of production.
effort from somewhere. The classical scholars discussed factor
intensities and factor intensity reversal in the factor proportion Today, many other production functions have emerged to take care
theory while making analysis on how countries can benefit of some shortcomings in the cobb-Douglass model. Prominent
from international trade. The classical scholars recognized the among them are the Leontief input-output model, Harold-Domar
importance of labour and capital in production because their model and constant elasticity of substitution (CES). The CES
argument is that the difference in value added by labour and capital model is a production function like the cobb-Douglass model
between two countries will form the basis for specialization in which allows any good to be produced with only two factor
production. However, they did not go further in quantifying the inputs. Moreover, it assumes CES between the two factor inputs.
relationship between input and output. The drawbacks of the CES model are two. The first is that the
model will break down when production of goods involves the
According Mishra (2007) the notion of inputs combination to use of more than two inputs, and number two, assumption of CES
produce a given output of goods is widely attributed to Knut between two factor inputs is difficult to obtain (Mishra, 2007;
Wicksell between 1900 and 1901. Mishra hints that Wicksell Uzawa, 1962).

International Journal of Economics and Financial Issues | Vol 8 • Issue 3 • 2018 143
 Amuka, et al.: Testing the Fit of Cobb-Douglass Production Function within Unrestricted Least Squares

The input-output model developed by Wassily Leontief is noted as Like every other theory, empirical tests have not produced uniform
the simplest of the production function developed to explain the result in the Cobb-Douglass case. Fraser (2002) tried to replicate
relationship between input and output. The production function Cobb-Douglass’s result using data of USA, Massachusetts, New
is based on factor proportion. In the Leontief’s model, when two South Wales, Victoria and New Zealand as before. He subjected
inputs that are not substitutable are combined in a fixed proportion the data to econometric tests and came out with a result that apart
to produce a given output of good, increasing one of the inputs and from New Zealand, result from USA, Massachusetts, New South
holding the other constant in the next round of production will not Wales and Victoria could hardly support laws of proportion as
change the level of output. For example, if 2 bags of sugar and 5 propounded by Cobb-Douglass. Study by Raval (2011) falls in
tons of flour are the right input combination to produce a given line with the Fraser’s finding. Raval in a micro study with data
quantity of bread, increasing the bags of sugar to 3 while the tons of from the manufacturing sector in America found that the capital
flour is unchanged will not increase the number of loaves of bread share in the cost of production varies from time to time and it is
produced. The model has been adjudged the simplest of all the models never constant as predicted by Cobb-Douglass model.
developed to explain the relationship between input and output.
Study by Duffy and Papageorgio (2000) equally rejects the validity
3. RELATED LITERATURE of the model as a good specification of aggregate production
function in a panel of 82 countries. Evidence from their panel
Cobb-Douglass production function shows different arrays of analysis in middle and low income countries points out that the
efficient production method facing a firm depending on the CES and the Cobb-Douglass production function may not be the
output level the firm wants to attain. The relationship it expresses correct aggregate production function. The finding in New Zealand
reveals the maximum amount of output possible from a particular by Szeto (2001) is similar to the result of Duffy and Papageorgio.
input combination. Arrow et al. (1961) did not support the notion Szeto discovered that there is substitution between primary factors
of constant returns to scale because factors of production are (value added) and import, and thus, rejected the Cobb-Douglass
never substituted on a constant proportion. Other arguments and specification. Other studies that invalidate the Cobb-Douglas
counter arguments have persisted on the validity of the aggregate model are Yusi (2016) in a study of the Pineapple production
production function as developed by Cobb-Douglass and others in in Southern Indonesia, and Hossain and Al-Amri (2010) in
(Felipe and Adams, 2005; Robinson, 1954; Hall, 1998). The the manufacturing sector of Oman. Both Yusi and Hossain and
argument of Robinson (1954) is that there is a mis-education in Al-Amri discover increasing return to scale since output more
production function that assumes all workers are alike. Making than double whenever input is doubled. In Romania, Silaghi and
the same line of argument, Hall (1998) contends that specifying Medesfalean (2014) reveal decreasing return to scale, as doubling
production function is always confronted with making a choice of of input leads to less than double in output.
the right algebraic form of argument to follow. In a study he carried
on crop yield, he identified that there is a cost to misspecification On the other hand, there are research evidences in support and
of production function when analysing the relationship between in validation of the Cobb-Douglass model. They include Chisasa
input and output. and Makina (2013) in South Africa, Adetunji et al. (2012) in
Nigeria and Ahmad and Khan (2015) in Pakistan. All these
Moreover, Felipe and McCombie (2001) noted the difficulty researches found validity in constant returns to scale property of
of aggregation of different kinds of commodities produced in a the Cobb‑Douglass production function. In Nigeria for instance,
country into a single output as assumed in Cobb-Douglass model. Adetunji et al. conclude that the reliance on the Cobb‑Douglass
Their argument is that production function is supposed to be in the model for policy will not hurt the economy. However, a
framework of microeconomy. In line with this argument, Guerrien contradicting result was found in Nigeria in a study by Abidemi
and Gun (2015) also doubt the possibility of aggregating quantities (2010). Abidemi discovered that as the banking sector doubles
of different kinds of commodities an economy can produce which its input use, its output more than doubles, suggesting increasing
will be called single output. Their position is that such assumption returns to scale in the banking sector in Nigeria.
is difficult to believe.
In the case of Nigeria, the work of Abidemi is a sectoral analysis
However, some scholars saw important contribution of which may be the reason for the display of increasing returns to
Cobb‑Douglass model in economic literature. Mishra (2007) scale. Private sector production is known to be associated with
stressed that Cobb-Douglass model has made useful contribution efficiency. On the other hand, the work of Adetunji is a macro
in the area of competitive equilibrium. Hong (2008) views Cobb- analysis but has a problem of specification. Adetunji et al. (2012)
Douglass model as a good description of production method. used capital expenditure of the government as the measure of the
Adetunji et al. (2012) accept the wide use of the production economy’s capital. The use of only government investment as a
function in works relating to productivity of factors and growth measure of the economy’s capital is grossly inadequate because
today. In a similar note, Biddle (2012) sees the Cobb-Douglas it will underestimate an economy’s capital and the contribution
model as very innovative for the reason that it shows that statistical of capital to total output. To overcome the problem, we used
method can be used to derive empirical relationship between input GFCF. Moreover, since the two studies used restricted least
and output. Moreover, Hagendorf (2013) accepts that despite the squares, we applied the unrestricted least squares so as to see if
fact that the capitalists are interested in profit, the Cobb-Douglass the Cobb‑Douglas model can still fit a macroeconomy data in the
production function can be used by the socialist planner. absence of restriction.

144 International Journal of Economics and Financial Issues | Vol 8 • Issue 3 • 2018
 Amuka, et al.: Testing the Fit of Cobb-Douglass Production Function within Unrestricted Least Squares

4. DATA AND METHODOLOGY et al. (2012) who used government capital investment, α and β
are coefficients to be estimated, ln = natural logarithm operator,
The Cobb-Douglas model was based on the assumption of constant and t = time horizon.
returns to scale, implying that in the production decision, whenever the
inputs used to produce a given output of goods is doubled, total output 5. UNRESTRICTED LEAST SQUARE
will automatically double. This restriction imposed in the model has
been the centre of controversy among economists (Bhanumurthy, Unrestricted least squares assume that the only existing
2002; Antras, 2004; Rana et al., 2010). In a study using America data, information is information within the parameters of the model.
Antras discovers that under restriction, the cobb-Douglas production It estimates the parameter of an equation without taking into
model fits America data. However, if the restriction is removed, account the linear equality restriction, that is, α+β = 1 (Adetunji
it cannot describe America economy well. Other criticism of the et al., 2012). Unrestricted least square uses the t-statistic instead
model include problem of specification (Hossain et al., 2012), and of the f-statistic. This involves conducting a test hypothesis after
identification (Mairesse, 2005). The criticisms of the Cobb-Douglas estimating α and β using the t-test, guided by the test statistic
model compelled Hossain et al. (2012) to make a modification of stated below:
the model in a study in the manufacturing sector in Bangladesh.
The Cobb-Douglas production model is important for developing (±+² ) −1
countries as the model can help answer the question of right input to=  (5)
var (α)+var(β)+2cov(α,β)
combination to obtain maximum output.
From 5,
We used quarterly data generated from Central Bank of Nigeria Ho = tcal<ttab, Cobb-Douglas model of constant returns to scale is
Statistical Bulletin (2012) between 1990 and 2012. High frequency validated
data is preferred in this case because production and investment H1 = tcal>ttab, Cobb-Douglas model of constant returns to scale is
is a daily affair and loss of information will be less in quarterly not validated.
data than the annual ones. That is, production decision is taken at
time interval less than a year (Zellner et al., 1966). 6. RESULTS
Restating the typical Cobb-Douglass production function in a 6.1. Unit Root Test
stochastic form, Test of stationarity of time series data is indispensable if the
research is to add value to literature since unstationary data leads
Y = ALαKβeui (2) to spurious result in econometric. The result of the unit root test
is presented in Table 1.
Where,
Y = Total output, Regression result presented in Table 2 can be presented in linear
L = Units of labour, form as modelled in equation (4). Therefore,
K = Units of capital, and α and β are elasticity of labour and capital,
and A is an efficiency parameter while e is the base of natural d-lnY = 0.003080 + 0.497442L + 0.077108K + et (6)
logarithm and ui is the stochastic disturbance term.
(Se) = (.002326) (0.062357) (0.017153)
From a priori expectation, one of the following conditions can
hold after estimation under the Cobb-Douglas restriction, viz.: (t) = (1.32) (7.98) (4.5)

α+β>1 Cov (α, β) = (0.3660)

α+β<1 By fitting Cobb-Douglass production in its exponential form,

equation 5 leads to:
α+β = 1
Y = 0.00308L0.4970 K0.0771 (7)
If the estimated function result is such that α+β = 1, Cobb-Douglass
hypothesis is validated, implying constant return to scale. From 7, under its restricted assumption,

Linearizing 2,
Table 1: Result of unit root test
lnYt = lnA+αlnLt+βlnKt+µt (3) Variable ADF test 5% level Order of
statistics integration
lnYt = λ+αlnLt+βlnKt+µt(lnA = λ) (4) d‑ln Y −16.171 −3.460 1 (1)
d‑ln L −11.957 −3.461 1 (1)
L is the labour force participation, K= gross fixed capital formation d‑ln K 11.176 −3.461 1 (1)
(GFCF) which is an important departure from the work of Adetunji Source: Authors’ computation

International Journal of Economics and Financial Issues | Vol 8 • Issue 3 • 2018 145
 Amuka, et al.: Testing the Fit of Cobb-Douglass Production Function within Unrestricted Least Squares

Table 2: Regression result (dependent variable d‑lnGDP)

Variable Coefficient Standard error t P> /t/
d‑lnL 0.497442 0.062357 7.98 0.0000
d‑lnK 0.077108 0.017153 4.5 0.0000
λ=C 0.003080 0.002326 1.32 0.1890
SE of regression 0.020382
Sum squared residual 0.036558
Log likelihood 226.6735
Mean dependent variable 0.013194
Standard deviation dependent variable 0.031512
Source: Authors’ computation. R2=0.59; Adjusted R2=0.58; DW=2.05; F‑stat=63.57; Prob F‑stat=0.0000

α+β = 0.497442+0.077108=0.575 of only capital investment of the public sector in the work of
Adetunji et al. (2012) was seen as inadequate to explain efficiency
Therefore, α+β<1, implying decreasing returns to scale. of capital in Nigeria. Government investment from experience is
less efficient in the developing countries compared to private ones.
However, since we are using the unrestricted approach, equation Unfortunately, despite important departures from the approach of
(5) applies. existing study, we have arrived at the same result that even if the
restriction imposed by the cobb-Douglass production is removed,
6.2. Examination of the Validity of Cobb-Douglass the model fits well in Nigeria economy. The result is consistent
Production Function within Unrestricted Least with the finding in Czech economy by Hajkova and Hurnik (2007).
Squares However, it is in opposite direction with the finding of Antras
Removal of the constraint imposed by the Cobb-Douglas model (2004) which shows that under its restricted form, Cobb‑Douglas
implies that our decision is not to be guided by the outcome of α+β. model fits American economy well and if the restriction is
Conclusion on the validity of the function will be based on the t-test of removed, it will no more fit the economy.
equation (4), and depends on whether tcal > tα/2 (n-k), where n = number
of observations, k = number of parameters estimated in the equation, Finally, research findings are trying to suggest that when it comes
and α is the 5% level of significance. From Table 2, tα/2 (n-k) = 2.0. to the use of Macro data, Cobb-Douglas model is valid, but with
the use of micro data, it will be invalidated (Raval, 2011; Abidemi,
Similarly, substituting the values in Table  2 into equation (5), 2010).
tcal = −0.4994. This is because the variance of α and β is the square
of their standard error.
REFERENCES
Since absolute value of t cal = 0.4994 and t α/2(n-k) = 2.0, Abidemi, A. (2010), Capital-labour substitution and banking sector
tcal<t  tab  (0.4994<2.0) in absolute term. It entails accepting the performance in Nigeria (1960-2008). Central Bank of Nigeria
null hypothesis of constant return to scale in Nigeria economy. Economic and Financial Review, 48(2), 109-130.
Therefore, the Cobb-Douglass production function of constant Adetunji, A., Ibraheem, A., Ademuyiwa, J.A. (2012), On restricted
returns to scale is valid for Nigeria economy. On the other hand, the least square: The cobb-douglass production function for Nigerian
Cobb-Douglass model fits well in Nigeria economy. For this reason, economy. Developing Countries Studies, 2(8), 68-73.
whenever inputs are doubled in production process, output will Antras, P. (2004), Is the U.S. aggregate production function cobb-douglas?
double as well in Nigeria. The finding falls in line with the earlier New estimates of the elasticity of substitution. Contributions to
study done in Nigeria by Adetunji et al. (2012) using restricted Macroeconomic, 4(1), 1-34.
lest squares, but different from the study by Abidemi. This has Arrow, K.J., Chenery, H.B., Minhas, B.S., Solow, R.M. (1961), Capital-
implication for economic policy and management in Nigeria. For labour substitution and economic efficiency. The Review of
Economics and Statistics, 47(3), 225-250.
one, when an analysis is made using aggregate economy data in a
Bhanumurthy, K.V. (2002), Arguing a case for the cobb-douglas
developing country like Nigeria, the public sector inefficiency will
production function. Review of Commerce Studies, 2002, 75-91.
outweigh private sector efficiency. Two, whenever an economy is Biddle, J. (2012), The Introduction of the cobb-douglas regression. Journal
public sector driven as in the case of Nigeria, increase in public of Economic Perspective, 26(2), 223-236.
sector investment without considerable increase in private sector Central Bank of Nigeria. (2012), Statistical Bulletin. CBN.
investment will not be enough to increase factor productivity. Chisasa, J., Makina, D. (2013), Bank credit and agricultural output in
South Africa: A cobb-douglass empirical analysis. International
Business and Economics Research Journal, 2(4), 387-398.
7. CONCLUSION
Duffy, J., Papageorgio, C. (2000), Cross-country empirical investigation
of aggregate production function specification. Journal of Economic
The study has gone through the rigours of unrestricted least squares Growth, 5, 87-120.
estimation procedure to test the fit of Cobb-Douglass production Felipe, J., Adams, F.G. (2005), ‘A theory of Production’ The estimation
function in Nigeria. Two important factors motivated the study of cobb-douglass function: A retrospective View. Eastern Economic
in Nigeria. One, Nigeria economy has continued to perform Journal, 31(3), 427-445.
poorly for more than two decades and a half despite continuous Felipe, J., McCombie, J.S.L. (2001), How Sound are the Foundations
investment by both the private and public sector. Two, the use of Aggregate Production Function? Economics Discussion Papers

146 International Journal of Economics and Financial Issues | Vol 8 • Issue 3 • 2018
 Amuka, et al.: Testing the Fit of Cobb-Douglass Production Function within Unrestricted Least Squares

Series No. 116. University of Otago. Journal of Economics, 107(4), 651-672.

Fraser, I. (2002), The cobb-douglass production function: A antipodean Miller, E. (2008), An Assessment of CES and Cobb-Douglass Production
defence? Part 1. Economic Issues, 7, 39-58. Functions. Available from: http://www.cbo.gov.
Guerrien, B., Gun, O. (2015), Putting an end to the aggregate function Mishra, S.K., editor. (2007), A Brief History of Production Function.
of production forever? Real-World Economic Review, 73(2015), Munich Personal RePEc Archives (MPRA, 5254).
99-109. Rana, M., Baten, A., Kamil, A.A. (2010), A stochastic frontier approach
Hagendorf, K. (2013), The Cobb-Douglass Production Function for empirical tests of efficiency wage models. Scientific Research
and Political Economy. Available from: http://www.ssrn.com/ and Essays, 5(11), 1234-1242.
abstract=1541468. Raval, D. (2011), Beyond cobb-douglass: Estimation of CES production
Hajkova, D., Hurnik, J. (2007), Cobb-douglas production function: The function with factor augmenting technology. Centre of Economic
case of a converging economy. Czech Journal of Economics and Studies, 2011, 5-11.
Finance, 57(9-10), 465-476. Robinson, J. (1953-1954), The production function and the theory of
Hall, H.H. (1998), Choosing an Empirical Production Function: Theory, capital. Review of Economics Studies, 21(2) 81-106.
Nonnested Hypotheses, Costs of Specifications, Agricultural Silaghi, M.P., Medesfalean, R. (2014), Some insights about determinants
Economics. Research Report No. 59. of economic growth in Romania: An empirical exercise. Theoretical
Hong, T.B. (2008), Cobb-douglas production function. In: Stewart, J., and Applied Economics, 21(6), 23-36.
editor. Calculus: Early Transcendentals. 6thed. United States, Szeto, K.L. (2001), An Econometric Analysis of a Production Function
Australia, Singapore, Canada, Australia, United Kingdom, Brazil: foe New Zealand, Treasury Working Paper No. 01/31.
Thomson Brooks/Cole/.p857-887. Uzawa, H. (1962), Production functions with constant elasticities of
Hossain, M., Majumder, A.K., Basak, T. (2012), An application of non- substitution. Review of Economic Studies, 29,291-299.
linear cobb-douglas production function to selected manufacturing Yusi, M.S. (2016), The analysis of production function and farm marketing
industries in Bangladesh. Open Journal of Statistics, 2, 460-468. efficiency of pineapple (Ananas comosus L Merr) in south Sumatera
Hossain, M.Z., Al-Amri, K.S. (2010), Use of Cobb-Douglas Production province, Indonesia. Journal of Agricultural Research, 11(23),
Model on Some Selected Manufacturing Industries in Oman. 1990-1998.
Available from: http//www.emeraldinsight.com/1753-7983. Zellner, A., Kmenta, J., Dreze, J. (1966), Specification and estimation
Mairesse, J. (2005), Panel-data estimates of the production function and of cobb-douglas production function models. Econometrica, 34(4),
the revenue function: What difference does it make? Scandinavian 784-795.

International Journal of Economics and Financial Issues | Vol 8 • Issue 3 • 2018 147