EKONOMETRIKA 1

INTRODUCTION

DEPARTEMEN EKONOMIKA DAN BISNIS

SEKOLAH VOKASI
 SYLLABUS
 This module talks about quantitative method
 To manage, to estimate, to infer and to narrate the data
 Make yourself enjoy with the module to motivate your study
 Main book:
 Wooldridge, J.M. 2016. Introductory Econometrics: A Modern Approach
 Supplementary:
 Gujarati, D.N. & Porter, D.C. 2008. Basic Econometrics 5th Ed.
 Greene, W.H. 2013. Econometric Analysis 7th Ed.
 SYLLABUS
 This module covers 6 topics:
 Linear regression (univariate and
multivariate)
 Qualitative regression
 Classical assumptions
 Time-series analysis
 Panel-data analysis
 Simultaneous regression
 Your tasks are to:
 Presence, follow and discuss
during the class and tutorial
 Submit assignment
 Submit projects for mid- and final
term
 SYLLABUS
 The individual assignment:
 Provide short analysis on
economic or business data
(distributed via Simaster)
 About 1000 words for the analysis
 Submit via Simaster by the
deadline
 The individual project:
 Provide a complete paper on
economic or business data analysis
 Introduction, Method, Analysis
and Conclusion
 About 5000 words in the body
 Descriptive and simple regression
analysis for mid-term
 Inference analysis for final term
 Submit via Simaster by the deadline.
 WHAT IS ECONOMETRICS?
 Econometrics  ‘Economic’ and
‘Metrics’
 A study to explore measurement in
economic context to help understand
situation better and in an objective way 
economics and metrics
 Serve as quantitative method through
extension of statistics
 We often call econometric model as
‘regression model’  towards to the
mean
THE NATURE OF ECONOMETRICS AND ECONOMIC DATA

 What is econometrics?
Econometrics is the use of statistical methods to analyze economic data.
Econometricians typically analyze nonexperimental data.

 Typical goals of econometric analysis:

Estimating relationships between economic variables.
Testing economic theories and hypotheses.
Evaluating and implementing government and business policy.

 Common applications
Forecasting macroeconomic variables (interest rates, inflation rates, GDP).
Forecasting non-macro variables (less visible).
THE NATURE OF ECONOMETRICS AND ECONOMIC DATA

 Steps in econometric analysis

1) Economic model (this step is often skipped)
2) Econometric model

 Economic models
 Maybe micro- or macromodels
 Often use optimizing behaviour, equilibrium
modeling, …
 Establish relationships between economic variables
 Examples: demand equations, pricing equations, …
THE NATURE OF ECONOMETRICS AND ECONOMIC DATA

 Econometric analysis requires data.

 There are several different kinds of economic data sets:

Cross-sectional data
Time series data
Pooled cross sections
Panel/Longitudinal data

 Econometric methods depend on the nature of the data used.

 Use of inappropriate methods may lead to misleading results.
THE NATURE OF ECONOMETRICS AND ECONOMIC DATA

 Cross-sectional data sets

 These may include samples of individuals, households, firms, cities, states, countries, or other units of interest at a given
point of time or in a given period.
 Cross-sectional observations are more or less independent.
 An example is pure random sampling from a population.
 Sometimes pure random sampling is violated, for example, people refuse to respond in surveys, or sampling may be
characterized by clustering.
 Cross-sectional data is typically encountered in applied microeconomics.
THE NATURE OF ECONOMETRICS AND ECONOMIC DATA
 Time series data
 This includes observations of a variable or several
variables over time.
 Examples include stock prices, money supply,
consumer price index, gross domestic product, annual
homicide rates, automobile sales, and so on.
 Time series observations are typically serially
correlated.
 Ordering of observations conveys important
information.
 Data frequency may include daily, weekly, monthly,
quarterly, annually, and so on.
 Typical features of time series include trends and
seasonality.
 Typical applications include applied macroeconomics
and finance.
THE NATURE OF ECONOMETRICS AND ECONOMIC DATA
Pooled cross sections
 Two or more cross sections are combined in one data set.
 Cross sections are drawn independently of each other.
 Pooled cross sections are often used to evaluate policy
changes.
 Example:
 Evaluating effect of change in property taxes on
house prices.
 Random sample of house prices for the year 1993.
 A new random sample of house prices for the year
1995.
 Compare before/after (1993: before reform, 1995:
after reform).
THE NATURE OF ECONOMETRICS AND ECONOMIC DATA
Panel or longitudinal data
 The same cross-sectional units are followed over time.
 Panel data have a cross-sectional and a time series
dimension.
 Panel data can be used to account for time-invariant
unobservables.
 Panel data can be used to model lagged responses.
 Example:
 City crime statistics; each city is observed in two
years.
 Time-invariant unobserved city characteristics may
be modeled.
 Effect of police on crime rates may exhibit time lag.
 ECONOMETRICS APPLICATION
 Individual consumption is expected to increase as with the increase in his/her
income
Is this hypothesis true?
How can this relationship be identified and modelled?
What are the underlying assumptions for this analysis?

 Simple model
 consumption (Y) is determined by the income (X)
Other factors outside the model is assumed to be constant:
 Data are difficult to find or unobservable
 We would need constant (a) to capture them
CAUSALITY AND THE NOTION OF CETERIS PARIBUS

 Consider this:

 Ceteris paribus: “other relevant factors being equal.”

 Most economic questions are ceteris paribus questions.
 It is important to define which causal effect one is interested in.
 It is useful to describe how an experiment would have to be designed to infer the
causal effect in question.
MODELLING IN ECONOMETRICS

 Econometric modelling is classified as

‘inferential statistics’
 Inferring conclusion by the probability

 Econometric model estimates the causal

effect given the dataset
^ =𝑎
𝑌 ^ + ^ 𝑋 +𝑢
𝑏  It needs to add the error component to
𝑖 𝑖 𝑖
accommodate the gap between actual and
estimated data..
 A good model requires a controlled error
 not necessarily need to be overfitting.
 MODELLING IN ECONOMETRICS

Create a
Estimating the Robustness
mathematical Inference
data check
relationship

1. Creating mathematical model

a. Use theory/hypothesis/argument to approximate the relationship
b. Assume a linearity/non-linearity of the model
c. Involve error term in the equation
2. Estimation
a. There are various estimation method in econometrics (OLS, MLE, GMM, etc.)
b. The estimation method should follow the data type and its business process
 MODELLING IN ECONOMETRICS

Create a
Estimating the Robustness
mathematical Inference
data check
relationship

3. Inference
a. Use a critical value {1%,5%,10%} as appropriate to test the hypothesis.
b. May use F-test or t-test or Z-test or Chi-sq test
4. Robustness check
a. Robustness check on the residual estimation given the estimation method.
b. Goodness-of-fit comparison for model selection (horse-run check)
SOME IMPORTANT THINGS IN ECONOMETRICS
 Model design determines the inferential purpose of the estimation
 Experiment, Quasi-Experiment and Non-Experiment
 Association vs Correlation vs Effect/Impact
 Model check should follow the estimation method
 Not every model will use classical assumptions
 Crucial is to control the error term
 Understand the design, data type, data collection and business process of the
data
 Robustness check and/or goodness-of-fit to validate the power of the model