Applied Econometrics


The teachers are Abel François and Rémi Generoso from USTL Lille 1 in Lille (FR).


Abel Francois Université de Lille – Sciences et Technologies
Bât. SH2, Cité scientifique
59655 Villeneuve d’Ascq Cedex

Rémi Generoso Université de Lille – Sciences et Technologies
Bât. SH2, Cité scientifique
59655 Villeneuve d’Ascq Cedex


The students are supposed to have a good knowledge of

  • Standard concepts of probability theory: Probability distributions, main asymptotic results
  • Basic theory of estimation and tests
  • Ordinary least squares estimation

Course contents

The basic linear model

  • Structure of the basic linear model

Identification and degrees of freedom

Hypotheses on the random term

  • Least Squares Estimation: OLS, GLS, FGLS

Endogenous variables

  • Endogenous variables.

Main reasons for endogeneity.

Problems raised by endogeneity.

  • Definition of instrumental variables. Strong and weak instruments
  • Estimation with instrumental variables: IV estimation
  • Tests

Testing endogeneity

Testing the quality of instruments

Time series analysis

  • General structure of stationary time series models:

Time series process, autoregressive and moving average processes,

Autocorrelation and partial autocorrelation.

Main types of econometric models with time series.

  • Non-stationarity: trends, random walks, unit roots
  • Univariate analysis of time series

Detection of non stationarity

Box and Jenkins analysis

Seasonal adjustment

  • Estimation of econometric models with stationary time series. VAR models.
  • Unit root time series and cointegration analysis

Univariate time series

Cointegrated time series: the Engle-Granger method and the Johansen method

Core reading

  • W.H. Greene, Econometric Analysis, Pearson Education, latest edition, ISBN 0-13-513740-3
  • G. S. Maddala, K. Lahiri, Introduction to Econometrics, Wiley, ISBN 978-0-470-01512-4
  • J.M. Wooldridge, Introductory Econometrics: A Modern Approach, Cengage Learning, ISBN 978-1-111-53104-1

Learning outcomes

Upon successful completion of the course, students should be able:

  • To use the main standard econometric methods for estimating an econometric model and testing hypotheses,
  • To prepare a dataset, choose the relevant estimation methods, detect autocorrelation and dealing with it, detect endogeneity and finding instruments.

At the end of the couse, students will have a more in-depth knowledge of time-series analysis and they are able:

  • To detect non stationarity,
  • To choose and estimate the relevant data generation process,
  • To carry out cointegration analysis.


The module consists of 12 lectures of 2 hours each. Students get one lecture per week.

Students mainly work with GRETL software, but are allowed to use other software as well. Students mainly do field work: Analyzing datasets, choosing a model, estimating and testing.


  • A practical exercise involving the use of an econometric package to analyse large data set and to write a report on the interpretation of the results and the associated inherent problems. This will be both word and time limited (normally one week). In assessing this exercise the examiners will expect to see (for a pass grade) the data correctly and efficiently organised and entered into a suitable computer package, a model clearly formulated and the parameters estimated by a suitable method, some analysis of the apparent success of the model as a framework of analysis, an interpretation suitable for non-specialist reader and a well organised report.
  • For a distinction grade the examiners would expect all of the above but also a clear indication that the student fully understood the procedures carried out and that no suitable standard procedure has been omitted, and, in addition, some extra feature which might be an imaginative individual approach for instance in terms of method, interpretation or link to the literature