1
Lecture 1 Introduction to Course: the Nature of Econometrics
1) The nature of econometrics
-in your introductory economics courses you learned about various theories
of economic behavior.
For instance, you learned in introductory macro that consumption is a function of
income.
Your professor may have used as an example something like the following:
c = 10 + .5y.
You may have wondered how the numbers 10 and 5 are found.
Well, the answer is by using econometrics!
Thus, econometrics is simply the attempt to discover quantifiable measures for the various theories
economists have developed.
Note an important point here, one which I will return to frequently in the
course.
We start with theory, and then try to quantify.
So theory comes first, and it is theory which tells us
the theoretical relationships we are interested in studying.
For example, it is theory that tells us that
consumption is a function of income.
2) Uses of econometrics
.
a) Describe economic reality
-an economic theory might say that consumption depends upon other
economic variables, let us say income and the rate of interest.
Because consumption depends on these
other variables, it is referred to as the dependent variable
.
The variables consumption depends upon, in this
case income and the rate of interest, are called independent variables
.
Let c = consumption, y = income, and
r = the interest rate.
We can write this economic theory as follows:
c =f(y,r).
When we do econometrics we attempt to find the true relationship between these variables. However, as we
will see, the information available to us may enable us to find only estimates of the true relationships.
We
might find the following:
c = 10 +0.5Y - 0.3r.
This equation would be giving us an estimation of the actual consumption function in this economy, i.e., it
would be describing the economic reality in this society.
In this last equation, the numbers 10, 0.5, - 0.3, are called
estimated regression coefficients
.
We will see
later why they are estimates.
For now we simply want to recognize that estimated regression coefficients
enable us to find the relationships between economic variables.
b) Once we find estimated regression coefficients, we may use them to test hypotheses
concerning
economic theories.
For instance, we know that economic theory tells us that consumption should be
positively related to income.
In our equation we see that the coefficient on income is positive, thus lending
credence to the theory.
If the estimated coefficient turned out to be equal to -.5, we would have doubts
about the hypothesized theory.
However, even when we find a value equal to .5 we still cannot be certain
that the hypothesized theory is verified.
This is because the coefficients we have found are estimated
coefficients.
Since they are estimates, and not necessarily the true values, we cannot be certain that the
theory has been substantiated.
But, from an intuitive pointed of view, we would expect that the further our
estimate is from 0, the more likely it is that our hypothesis is true.
Note that if the coefficient were in fact
equal to 0, then there would be no theoretical relationship between consumption and income.