2
3
1. GOALS AND OBJECTIVES OF THE DEVELOPMENT OF THE DISCIPLINE
1.1. Objectives: the main objective of the course "Econometrics (advanced level)" is to
deepen the Students' understanding of the theoretical foundations of modern econometric data analysis
methods and to teach the correct use of tools in practice when working with specialized econometric
software.
1.2. Objectives: learn to carry out the collection, processing, analysis and systematization of
information on research, choice of methods and means of solving research objectives; to develop
theoretical and econometric models of the processes, phenomena and objects relating to the sphere of
professional activity, to evaluate and interpret the results; predict the dynamics of the main socioeconomic indicators of the enterprise, industry, region and economy.
2. PLACE OF DISCIPLINE IN THE STRUCTURE OF THE EDUCATIONAL
PROGRAM
2.1. Cycle (block) OP: B1.B.
2.2. Communication with other disciplines of the curriculum
A list of previous disciplines
1 .List follow disciplines, types of work
1. Microeconomics (advanced level)
1. External International Business Environment
2. Macroeconomics (advanced level)
in the EU
2. Features of the international commercial
level)level)Foreign language of
operations in Russia
professional
3. REQUIREMENTS FOR DEVELOPMENT RESULTS OF THE DISCIPLINE
Code
GC-1
PC-8
PC-9
Formed competencies
Name
Knowledge, skill, master
General cultural competencies (GC)
Know the basics of mathematical proof
To be able to apply the methods of analysis and
The ability to abstract
synthesis in the subject area
thinking, analysis, synthesis
Master methodology to improve knowledge in the field
of econometrics
Professional competencies (PC)
Analytical work
Know the capabilities of modern information
Ability prepare analytical
technologies in the field of econometrics
materials for evaluation
activities in the field of
To be able to apply econometric methods to solve
economic policy and strategic applied problems; self-study econometric methods to
decision-making at the micro
solve problems in the professional sphere
and macro levels
Master applied econometric methods of data analysis
To know the modern methods of econometric analysis;
The ability to analyze and use possible restrictions on the use of econometric methods
various sources of information Be able to use modern software solutions for economic
for economic calculations
and statistical and econometric problems
Master the techniques of applied econometric analysis;
4
modern packages of applied econometric software
Know the theoretical material on the basics of
The ability to forecast the main econometrics
socio-economic indicators of
To be able to interpret the results of econometric
the enterprise, industry, region modeling
and economy as a whole
Master the techniques of applied econometric analysis;
modern packages of applied econometric software
PC-10
4. STRUCTURE AND CONTENT OF THE DISCIPLINE
16
8
6
3
2
1
Classroom classes - full-time course
View class, module, topic and summary
competences
including
interactive hours
number of hours
4.1.
Laboratory works
Module 1 «The linear regression model and its specification.
Discrete choice models»
Topic 1.1 «The Classical Multiple Linear Regression Model»
The linear regression model. The ordinary least-squares method (OLS).
Extended least squares method (ELS). Dummy variables. Restricted
least squares. The maximum-likelihood method.
PC-8
PC-9
PC-10
Topic 1.2 «Model specification in regression analysis»
2
1
2
1
10
5
Multicollinearity and elimination methods. Specification errors and
their detection. Choosing the optimal set of regressors and functional
forms of the regression dependence. Heteroskedasticity, its causes and
methods of detection and elimination. Weighted least squares (WLS).
Residual autocorrelation, its causes, detected and elimination methods
Endogeneity of variables. The case of correlated regressors and random
error. Errors of measurement variables. Instrumental variables. The
Hausman test.
Topic 1.3 «Logit and Probit models»
Discrete dependent variables: nominal, ranked and quantitative.
Binary choice models. Probit and Logit models.
Interpretation of
coefficients in binary choice models. The maximum-likelihood method
in Probit and Logit models . Goodness-of-fit testing for models.
Module 2 «Time series models and panel data models»
Topic 2.1 «Time series models»
GC-1
PC-8
PC-9
PC-10
PC-9
PC-10
4
1
Box-Jenkins models. Distributed lag models (partial adjustment model,
adaptive expectations model). The Granger causality test.
GC-1
PC-8
PC-9
PC-10
2
1
Topic 2.2 «Non-stationary time series»
GC-1
5
Imaginary regression. The unit root. Unit root tests. Cointegration of
time series. Error correction model.
Topic 2.3 «Panel data»
2
1
Advantages of using panel data. Difficulties in working with panel data.
Error components model. Model specification. Fixed and random
individual effects.
Topic 2.4 «Models for panel data»
2
2
Operators «Between» and «Within». Types of estimates. Comparative
analysis of estimates. Specification tests for panel data models. The
Hausman test. Testing fixed effects. Testing random effects.
16
8
6
3
Practical lessons
Module 1 «The linear regression model and its specification.
Discrete choice models»
Topic 1.1 «The Classical Multiple Linear Regression Model»
2
1
The linear regression model. The ordinary least-squares method (OLS).
Extended least squares method (ELS). Dummy variables. Restricted
least squares. The maximum-likelihood method.
PC-8
PC-9
PC-10
GC-1
PC-8
PC-9
PC-10
GC-1
PC-8
PC-9
PC-10
PC-8
PC-9
PC-10
Topic 1.2 «Model specification in regression analysis»
2
1
2
1
10
5
4
1
Multicollinearity and elimination methods. Specification errors and
their detection. Choosing the optimal set of regressors and functional
forms of the regression dependence. Heteroskedasticity, its causes and
methods of detection and elimination. Weighted least squares (WLS).
Residual autocorrelation, its causes, detected and elimination methods
Endogeneity of variables. The case of correlated regressors and random
error. Errors of measurement variables. Instrumental variables. The
Hausman test.
Topic 1.3 «Logit and Probit models»
Discrete dependent variables: nominal, ranked and quantitative.
Binary choice models. Probit and Logit models.
Interpretation of
coefficients in binary choice models. The maximum-likelihood method
in Probit and Logit models . Goodness-of-fit testing for models.
Module 2 «Time series models and panel data models»
Topic 2.1 «Time series models»
Box-Jenkins models. Distributed lag models (partial adjustment model,
adaptive expectations model). The Granger causality test.
Topic 2.2 «Non-stationary time series»
2
1
Imaginary regression. The unit root. Unit root tests. Cointegration of
time series. Error correction model.
6
GC-1
PC-8
PC-9
PC-10
PC-9
PC-10
GC-1
PC-8
PC-9
PC-10
GC-1
PC-8
PC-9
PC-10
Topic 2.3 «Panel data»
2
1
2
Operators «Between» and «Within». Types of estimates. Comparative
analysis of estimates. Specification tests for panel data models. The
Hausman test. Testing fixed effects. Testing random effects.
Topics, sections for an independent training: themes of essays, tests,
recommendations for the use of literature, computers, and others.
30
Topics, sections for self-training
2
Classical multiple linear regression model
2
Regression model specification.
4
Logit and Probit models
6
Time series models. Non-stationary time series
4
Models for panel data
12
Simultaneous equations models
10
GC-1
PC-8
PC-9
PC-10
Independent work of students - full-time training
number of
hours
4.2.
Advantages of using panel data. Difficulties in working with panel data.
Error components model. Model specification. Fixed and random
individual effects.
Topic 2.4 «Models for panel data»
Topics and questions for discussion
Вариант задания:
Write a review of the presented article. 1) Did you find this article
7
competences
2
GC-1
PC-8
PC-9
PC-10
GC-1
PC-8
PC-9
PC-10
GC-1
PC-8
PC-9
PC-10
GC-1
PC-8
PC-9
PC-10
GC-1
PC-8
PC-9
PC-10
GC-1
PC-8
PC-9
PC-10
GC-1
PC-8
PC-9
PC-10
GC-1
PC-8
PC-9
40
interesting issues are discussed? 2) How much proved the authors answers
to these questions? 3) Whether the authors are familiar with the literature
on this subject? 4) Does the theoretical analysis is carried out in this
article and whether the original theory? 5) Is there an empirical analysis
and whether the original empirical analysis? 6) Do the authors use the
appropriate data? 7) How should the authors to improve the analysis?
Total labor input of independent work (hours)
PC-10
5. ASSESSMENT TOOLS FUND
5.1.
Fund assessment tools for monitoring
Module 1. The linear regression model and its specification. Discrete choice models.
Questions
1.
Multiple linear regression model.
2.
Ordinary least-squares method (OLS)
3.
Extended least squares method (ELS).
4.
Dummy variables..
5.
Restricted least squares.
6.
The maximum-likelihood method.
7.
Multicollinearity and elimination methods.
8.
Specification errors and their detection.
9.
Choosing the optimal set of regressors and functional forms of the regression
dependence
10.
Heteroskedasticity, its causes and methods of detection and elimination. Weighted least
squares (WLS).
11.
Residual autocorrelation, its causes, detected and elimination methods.
12.
Endogeneity of variables. The case of correlated regressors and random error.
13.
Errors of measurement variables. Instrumental variables. The Hausman test..
14.
Discrete dependent variables: nominal, ranked and quantitative. Binary choice models.
Probit and Logit models.
15.
Interpretation of coefficients in binary choice models. The maximum-likelihood
method in Probit and Logit models. Goodness-of-fit testing for models.
Multilevel tasks
1.
*
1, y     x    0
Consider the binary choice models y  
*
0, y     x    0
y\x
0
1
0
n00
n01
1
n10
n11
a)
Prove that the model can not be solved if n10=0 for any error distribution function
b)
Estimate the coefficients of logit-model P  y  1      x
c)
Prove that for the logit-model P̂  y  1 
F  z  : f  z   0

8
n10  n11
n10  n11  n01  n00

d)
Estimate LP-model (ε~Uniform[-1, 1]). What is P̂  y  1
2.
Suppose you have estimated probit-model and have found that its coefficients are proportional
to the coefficients of LP-model (ε ~ Uniform [-1, 1]). What approximately the proportionality coefficient
should be equal?
3.
1 , y*  1

Consider tobit-model: y   y* , 1  y*   2 , y*  xβ   ,

*
 2 , y   2
where mistakes have distribution density f(z)
a)
Find the distribution y
b)
Find the log-likelihood function for estimating the vector β
E y
c)
Find
x
Laboratory works
Description of the data files
1) File MROZ contains 753 observations of random sample married women in the USA (1975), the
first 428 of which work, and the remaining 325 are not working.
The variables (for women, unless otherwise):
LFP - a dummy variable equal to 1 if the woman worked in 1975, 0 otherwise;
WHRS - number of hours of work in 1975;
KL6 – the number of children under 6 years old in the household;
K618 - the number of children from 6 to 18 years old in the household;
WA - age;
WE - education in years;
WW -average hourly earnings in 1975 (in US dollars);
RPWG - wages, reported during a survey in 1976 (not the same WW).
To use the sub-sample with the variable, you must select the employees in 1975, for which the LFP = 1;
HHRS - the number of hours of her husband in 1975;
HA - age of the husband;
HE - Education husband in years;
HW - husband's salary in 1975;
FAMINC - family income. The variable used to construct variable unearned income;
MTR - marginal rate of tax per cent;
WMED - education in the mother's;
WFED - education of his father in years;
UN - the unemployment rate in the State of residence, in%;
CIT dummy variable equal to 1 if the woman lives in a big city, otherwise 0;
AX - the number of years of experience.
Exercise 1: Reviewing the Evidence
a) Using the data from Mroz, calculate the expectation and the standard deviation, minimum and
maximum for all 19 variables. Comment on the results.
b) Run a) separately for working women (428 cases) and not working (325 observations). Comment on
the results. Do subsamples differ in the variables WA, WE, K618, HA, HE, HHRS? What about the descriptive
statistics for the explanatory variables on KL6 and HW? Interpret the difference in AX.
c) Construct a variable PRIN unearned income by rule PRIN=FAMINC(WHRS∙WW). Calculate
descriptive statistics.
d) For the subsample of 428 women employed to compute the value of a new variable LWW = LN
(WW). For the entire sample, build variables AX2 = AX ∙ AX, WA2 = WA ∙ WA. Construct Mincer
regression model by constant, WA, WE, CIT, AX and AX2 only for working women. Comment on the results.
9
Using the estimates of the regression parameters, calculate the predicted values of the logarithm of wages for
unemployed women, and name the variable FLWW. Compare the average values of the variables LWW and
FLWW. Interpret the results. Create a new variable LWW1, LWW equal to employees, and equal FLWW for
not working. Calculate the mean and standard deviation You should get 1.10432 and 0.58268 respectively.
Exercise 2: Evaluation of labor supply equation (procedure I)
a) Check that the WHRS = 0, = 0 when the LFP. Build OLS regression for the whole sample WHRS the
constant WHRS KL6, K618, WA, WE, LWW1 and PRIN. The signs of the coefficients correspond to economic
theory? If not - why? What is R2? Why is it not enough?
b)The effect of uncompensated wage by the number of hours of work by Mroz can be calculated as
Hi/Wi=a1/Wi and the income effect - as Hi/Vi=a2. Appropriate elasticity calculated as lnHi/lnWi=a1/Hi and
lnHi/lnVi=a2Vi/Hi, where a1 and a2 – regression coefficients for the variables lnWi and Vi respectively.
Calculate the elasticity of supply hours for the regression a) of wages and unearned income. Elasticity of of
wages compensated or not? Why? Calculate the effect of uncompensated wage by the number of hours of work
and the income effect on Mroz. Give an interpretation of the results.
c) Explain principal flaws of used in a) labor supply estimation procedure.
The structure and content of the fund assessment tools are presented in Appendix 1 to the
working program of discipline.
5.2.
Fund assessment tools for intermediate certification in the form of credit
Credit Questions:
1.
Multiple linear regression model. Ordinary least-squares method (OLS). Extended least
squares method (ELS).
2.
Dummy variables. Restricted least squares.
3.
The maximum-likelihood method.
4.
Multicollinearity and elimination methods.
5.
Specification errors and their detection. Choosing the optimal set of regressors and functional
forms of the regression dependence.
6.
Heteroskedasticity, its causes and methods of detection and elimination. Weighted least
squares (WLS).
7.
Residual autocorrelation, its causes, detected and elimination methods.
8.
Endogeneity of variables. The case of correlated regressors and random error. Errors of
measurement variables. Instrumental variables. The Hausman test.
9.
Discrete dependent variables: nominal, ranked and quantitative. Binary choice models. Probit
and Logit models. Interpretation of coefficients in binary choice models.
10.
The maximum-likelihood method in Probit and Logit models. Goodness-of-fit testing for
models.
11.
Box-Jenkins models.
12.
Distributed lag models (partial adjustment model, adaptive expectations model).
13.
The Granger causality test
14. Imaginary regression. The unit root. Unit root tests.
15. Cointegration of time series. Error correction model.
16. Advantages of using panel data. Difficulties in working with panel data.
17. Error components mode. Model specification. Fixed and random individual effects.
18. Operators «Between» and «Within». Types of estimates. Comparative analysis of estimates.
Specification tests for panel data models. The Hausman test. Testing random effects. Testing fixed
effects.
19. The concept of systems of equations that are used in econometrics. Structural and reduced form models. The
system of simultaneous equations. Endogeneity and causality.
10
20. Problems of identification. Conditions of order and rank. Estimation methods. Recursive systems. Indirect
OLS. The two-step method of OLS and instrumental variables method.
The three-step OLS. Dynamic systems.
6. TRAINING METHODICAL AND INFORMATION MAINTENANCE OF THE
DISCIPLINE
6.1. Core books and further reading
№
1
2
1
2
3
1
№
1
References
Core books
Econometrics: Textbook / ed. I.I.Eliseeva. - М.: Prospect, 2014.
50
Putko B.A. Econometrics: Textbook / B.A. Putko, N.S. Kremer; ed. N.S. KremerPublisher: Unity-Dana, 2012.;
Idem [Electronic resource]. - URL:
http://biblioclub.ru/index.php?page=book&id=118251
Additional literature
Econometrics. Initial course: Textbook / J.Magnus, P.Katyshev, A.Peresetsky. М.: Delo, 2005.
Unlimited
access
for
registered
users
Ayvazian S. A. Applied Statistics and Econometrics Basics / S. A. Ayvazyan, V.
S. Mkhitaryan. – M. : Unity, 1998.
Eliseeva I.I. Econometrics: Textbook / I.I.Eliseeva, S.V. Kurisheva, T.V.
Kosteeve. - М.: Finance and statistics, 2005.
Idem [Electronic resource]. - URL:
http://biblioclub.ru/index.php?page=book&id=260409
Methodical works
Arzenovsky S.V., Toropova T.V. Econometrics in Eviews. Practicum. Rostov-onDon: RSUE, 2010.
48
1
Unlimited
access
for
registered
users
10
6.2. The list of resources information and telecommunication network "Internet"
output
The journal "Applied Econometrics". Electronic resource http://appliedeconometrics.cemi.rssi.ru/
6.3.
№
1.
2.
3.
Number of
copies
List of software
output
Econometric Views 6.0
Statistica 6.0
MS Excel
6.4. The list of information systems
Name: Information Systems
11
1.
2.
3.
Databases Rosstat:
http://www.gks.ru/wps/wcm/connect/rosstat_main/rosstat/ru/statistics/databases/
Statistics of the Central Bank of the Russian Federation: http://www.cbr.ru/statistics/
Internet resources www.econ.kuleuven.ac.be/gme www.statsoft.ru
6.
LOGISTICS OF DISCIPLINE
Premises for all types of work, provided for by the curriculum, equipped with the necessary
specialized training furniture and technical means of education, display equipment used for lectures.
Practical exercises are conducted in computer classes, jobs that are equipped with the licensed
software and Internet access (a. 513, 516).
7. GUIDELINES FOR THE DEVELOPMENT OF DISCIPLINE
Guidelines for the development of the discipline of "Econometrics (advanced level)" addressed
to undergraduates of full-time training.
The curriculum in the direction of preparation "Economy" provides the following types of
activities:
- laboratory works;
- practical lessons.
In laboratory studies skills in using econometric methods to solve specific problems are
developed.
In preparation for the laboratory work the student must:
- examine the recommended textbooks;
- to prepare the answers to all the questions on the topic being studied;
- do homework, the teacher recommended when studying each topic.
In preparation for laboratory work, students can use the consultations of the teacher.
During the practical exercises deepened and consolidated knowledge on a number of issues
discussed in the lectures.
In preparation for the practical training every undergraduate should:
- Examine the recommended textbooks;
- study lecture notes;
- prepare the answers to all the questions on the topic being studied.
In consultation with the teacher to prepare undergraduate essay, report or report on employment.
Issues not discussed in lectures and practical classes, must be examined in the course of selfstudy. Control of independent work of students on the curriculum of the course is carried out in the
course of employment by means of oral questioning or testing. In the course of independent work each
student is required to read the primary and secondary literature on the subject under study, lecture
notes complement the missing material, excerpts from primary sources recommended. Highlight
incomprehensible terms find their value in professional or encyclopedic dictionaries.
With the implementation of various types of study used a variety (including online) teaching
methods, in particular:
- Interactive board for preparing and conducting lectures and seminars;
To prepare for, monitoring and interim assessment of students can use e
http://library.rsue.ru/ university library.
Also, students can take home the necessary books on the subscription or use the university
library reading rooms of the university.
12