INTRODUCTION TO
ECONOMETRICS
Truong Dang Thuy truong@dangthuy.net
Econometrics…
…is the application of mathematics and statistical methods to
investigate economic relations. It is:
▪ the quantitative analysis
▪ of observational data
▪ guided by economic theories
USES OF ECONOMETRICS
▪ Uses:
▪ assessing economic theories
▪ analysing economic data
▪ and forecast
▪ In all cases: it involves the investigation of the relationship
between two (or more) variables
A SIMPLE ILLUSTRATION
REAL GDP (BIL. VND)
180000
160000 One more thousand of
140000
laborers result in 39 bil. VND
increase in real GDP
120000
100000 y = 39x - 10495
80000
60000
40000
20000
0
0 500 1000 1500 2000 2500 3000 3500 4000 4500
LABOR FORCE (1000 PEOPLE)
RELATIONSHIP BETWEEN VARIABLES: EXAMPLES
▪ Individual: Does a master degree improve my wage?
▪ Need to analyse the impact of education on wage.
▪ Producer: How much my output increase if I hire more labor?
▪ Input and output (production function)
▪ Seller: How much would the demand for my product increase if I reduce the price
by $X?
▪ Price and quantity demanded (demand function)
▪ Employer: Is laborer working more with higher wage?
▪ Wage and working hours (individual labor supply function)
CETERIS PARIBUS
We did some simple tests above (t-test and chi2 test)
We tried to answer the research questions – the relationship between two variables.
But what if the two groups are different in many other aspects?
CETERIS PARIBUS
▪ For example, we found that women earn 98K/day and men 102K/day.
▪ A t-test rejected the difference.
▪ So, can we conclude that there is no gender discrimination?
CETERIS PARIBUS
▪ What if female in the sample are more skilful (and thus work more efficiently) than
men?
▪ In that case, the no-gender-difference we found has not been separated from skills.
▪ We may want to know: are women and men of the same skill level being paid the
same rate?
▪ Or more accurately: are women and men of the same everything being paid the
same rate?
▪ When we separate the effects of all other factors (i.e. holding other things equal),
the remained difference is the gender discrimination.
▪ How to do that?
ECONOMETRIC TECHNIQUES
Multiple regressions
Advanced methods to obtain unbiased estimates
MULTIPLE REGRESSION
▪ Multiple regression function:
𝑤𝑎𝑔𝑒 = 𝛽0 + 𝛽1 𝑔𝑒𝑛𝑑𝑒𝑟 + 𝛽2 𝑒𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒 + 𝛽3 𝑋
▪ 𝑤𝑎𝑔𝑒 is the dependent variable
▪ 𝑔𝑒𝑛𝑑𝑒𝑟 and 𝑒𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒 are independent variables
▪ in this equation, 𝛽1 indicates the gender difference of wage, holding experience
level equal.
▪ By adding 𝑒𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒, we control for the experience effect.
▪ We can control other effects by adding more independent variables.
UNBIASED ESTIMATES
𝑦 = 𝛽1 𝑋1 + 𝛽2 𝑋2
▪ What other independent variables should we include? Many.
▪ Do we miss one or more important variables? Test.
▪ Are we including an irrelevant variable? Test.
▪ Is there any problem that make our estimates of 𝛽 biased? Test.
▪…
Many econometric techniques to obtain unbiased estimates.
THEORIES GUIDE
ECONOMETRIC ANALYSES
So independent variables are selected based on economic theories.
Let’s go through some examples…
MODEL SPECIFICATION:
WAGE FUNCTION
▪ Question: Does a master degree improve my wage?
▪ The dep var is wage
▪ What are the independent variables (regressors)?
▪ Mincer (1974) wage function:
▪ Schooling
▪ Experience
▪ Other individual characteristics
▪ Should also look at other empirical studies to search for case-specific regressors.
MODEL SPECIFICATION:
PRODUCTION FUNCTION
▪ Question: How much my output increase if I hire more labor?
▪ The dep var is output. And thus we estimate the production function.
▪ By theory, production function shows the relationship between inputs and output.
▪ So all the inputs should be regressors.
MODEL SPECIFICATION:
DEMAND FUNCTION
▪ Question: How much would the demand for my product increase if I reduce the
price by $X?
▪ The dep var is quantity demanded.
▪ By demand theory, regressors include
▪ Own price
▪ Prices of related commodities
▪ Income
▪ Other variables reflecting preferences
MODEL SPECIFICATION:
LABOR SUPPLY FUNCTION
▪ Question: Does subsidies (cash transfers) to the poor reduce their working efforts?
▪ Dep var: working time
▪ Regressors (by theory of individual labor supply:
▪ Wage
▪ Variables reflecting preferences for leisure
INTRODUCTION TO
THE COURSE
So you need to study appropriate techniques to obtain unbiased
coefficients…
That’s why you take this course.
LECTURES
▪ Lecture 1: Introduction to Econometrics
▪ Lecture 2: R Basics
▪ Lecture 3: Linear Regression Model
▪ Lecture 4: Functional forms
▪ Lecture 5: Multicollinearity
▪ Lecture 6: Heteroskedasticity
▪ Lecture 7: Panel data models
▪ Lecture 8: Binary response models
EXPECTED LEARNING OUTCOMES
▪ Know the econometric models introduced in this course
▪ what they are,
▪ in which situations they are appropriate for application
▪ Estimate the models in RStudio
▪ Interpret the estimation results
ASSIGMENTS
▪ For this course, you will complete 10 assignments
▪ 24 hours to complete, you can choose when to start within a time range
▪ each has 20 multiple choice questions
▪ score is based on the number of correct answers
▪ the first 10 correct answers earn no score
▪ the next 10 correct answers earn 10 scores each
THE PLATFORM
▪ you can do assignments here https://kinhteluong.online/student/login.php
▪ the first time logging in, please click reset password, enter your UEH email, check
your email (sometimes in spam folder), click the link provided to reset your
password.
▪ then you can log in. Remember that username of your UEH student ID, not the email
▪ At this plaform, you can
▪ download slides and data
▪ read materials
▪ practice with multiple choice question for each skills
▪ do the assignments
▪ check the assignment scores
▪ check your progress
