DEMAND ESTIMATION:

THE WONDER OF ECONOMETRIC ANALYSIS

Why do we have to estimate demand?
because demand dictates the music.
What is Econometric Analysis?
Econometrics means economic measurement.
Consists of the application of mathematical statistics to eco
nomic data to lend empirical support to the models constructe
d by mathematical economics and to obtain numerical results
(Gerhard 1968).
Econometrics may be defined as the social science in whic
h the tools of economic theory, mathematics, and statistical in
ference are applied to the analysis of economic phenomena
(Goldberger 1964).
Econometrics is concerned with the empirical determinatio
n of economic laws (Theil 1971).
Why is it a separate discipline?
Economic theory makes statements or hypotheses that are
mostly qualitative in nature (e.g. the law of demand). The law
does not provide any numerical measure of the relationship.
The main concern of mathematical economics is to express eco
nomic theory in mathematical form without regard to measurabilit
y or empirical verification of the theory.
Economic statistics is mainly concerned with collecting, processi
ng, and presenting economic data in the form of charts and table
s. It does not go any further. The one who does that is the econo
metrician.
METHODS OF ESTIMATING DEMAND
Consumer surveys (e.g. interviews, surveys)
Pros: up-to-date information, generally easy to conduct
Cons: validity, reliability and sample bias
Market experiments (e.g. lab and field experiments)
Pros: direct observation
Cons: lack of control, lesser sample, length of time needed
Statistical methods (e.g. econometric analysis)
Pros: more accurate and reliable
Cons: require big data, complex computations
SEVEN STAGES OF ECONOMETRIC ANALYSIS
1. Statement of theory and hypotheses
2. Model specification
3. Data collection
4. Estimation of variables
5. Checking goodness of fit
6. Hypothesis testing
7. Forecasting
1. STATEMENT OF THEORY AND HYPOTHESIS

Hypothesis - a supposition or proposed explanation

made on the basis of limited evidence as a starting
point for further investigation.
Theory - a supposition or a system of ideas intende
d to explain something, especially one based on ge
neral principles independent of the thing to be explai
ned.
1. STATEMENT OF THEORY AND HYPOTHESIS
Example of a theory:
JM Keynes states that on average, consumers increase their
consumption as their income increases, but not as much as the
increase in their income (i.e. marginal propensity to consume)
In mathematical terms:

Where MPC is the marginal propensity to consume

2. MODEL SPECIFICATION
Deterministic vs. Statistical
Deterministic relationship relationships are k
nown with certainty
(e.g. Revenue = Price x Quantity sold, or R = PQ)
Statistical relationship involves an element o
f uncertainty
(e.g. Quantity = function of Price, or Q = f(P))
2. MODEL SPECIFICATION
Example:
Single-equation model for consumption:
Y = 1 + 2X 0< <1 2

Y = consumption expenditure (dependent variable)

X = income (independent, or explanatory variable)
1 = the intercept
2 = the slope coefficient

The slope coefficient 2 measures the MPC.

Graphically
2. MODEL SPECIFICATION
Note: The relationships between economic variabl
es are generally inexact. In addition to income, oth
er variables affect consumption expenditure. For e
xample, size of family, ages of the members in the
family, family religion, etc., are likely to exert some
influence on consumption.
2. MODEL SPECIFICATION
To allow for the inexact relationships between ec
onomic variables, (I.3.1) is modified as follows:
Y = 1 + 2X + u

where u, known as the disturbance, or error, term, is a random variable

that has well-defined probabilistic properties. The disturbance term u m
ay well represent all those factors that affect consumption but are not t
aken into account explicitly.
Graphically
3. DATA COLLECTION
TYPES OF DATA
Time-series data single entity at different period
s
Cross-section data different entities at single pe
riod of time
SOURCES OF DATA
Records of firms
Commercial and private agencies
Official sources (i.e. public agencies)
3. DATA COLLECTION

Y = consumption
expenditure

X = gross domestic
product (which in
economics is a good
measure of national
income)
4. ESTIMATION OF THE ECONOMETRIC MODEL
Regression analysis (e.g. least squares method) is th
e main tool used to obtain the estimates. Using this tech
nique and the data given in Table I.1, we obtain the follo
wing estimates of 1 and 2, namely, 184.08 and 0.706
4. Thus, the estimated consumption function is:

Y = 184.08 + 0.7064X
5. CHECKING GOODNESS OF FIT
5. CHECKING GOODNESS OF FIT
(Continuation of example) The regression line fit
s the data quite well*. The slope coefficient (i.
e., the MPC) was about 0.70, an increase in real
income of 1 dollar led, on average, to an increas
e of about 70 cents in real consumption.
*this may be validated through the use of a statistical techn
ique known as the coefficient of determination (commonly k
nown as R squared)
6. HYPOTHESIS TESTING
Keynes expected the MPC to be positive but less than 1.
In our example we found the MPC to be about 0.70. But
before we accept this finding as confirmation of the theor
y, we must first ask, is 0.70 statistically less than 1? If it i
s, it may support Keynes theory.
Such confirmation or refutation of economic theories on t
he basis of sample evidence is based on a branch of sta
tistical theory known as statistical inference (hypothes
is testing).
7. FORECASTING
To illustrate, suppose we want to predict the mean consumption expen
diture for 1997. The GDP value for 1997 was 7269.8 billion. Using the
model, dollars consumption would be:

Y1997 = 184.0779 + 0.7064 (7269.8) = 4951.3

The actual value of the consumption expenditure reported in 1997 wa
s 4913.5 billion dollars. The estimated model thus over-predicted th
e actual consumption expenditure by about 37.82 billion dollars. We c
ould say the forecast error is about 37.8 billion dollars, which is about 0
.76 percent of the actual GDP value for 1997 (not bad, right?).
8. USING THE MODEL TO PROPOSE ECONOMIC POLICIES
Suppose we have the estimated consumption function given in Suppos
e further the government believes that consumer expenditure of about
4900 will keep the unemployment rate at its current level of about 4.
2%. What level of income will guarantee the target amount of consump
tion expenditure? If the regression results given in seem reasonable, si
mple arithmetic will show that:
4900 = 184.0779 + 0.7064X

which gives X = 7197, approximately. That is, an income level of a

bout 7197 (billion) dollars.
CONCLUSION

Economic concepts and theories, when com

bined with data collection, statistics and pre
dictive analytics, is undoubtedly a game-ch
anger both in the macro- and micro-levels.