The IARIW is offering two concurrent training courses immediately preceding the General Conference on the topics of national accounts and non and semi-parametric methods for data classification and categorization. Details on the sessions are found below. The sessions are open to individual IARIW members as well as employees of IARIW institutional members at no charge. As space may be limited, persons interested are encouraged to register as soon as possible. The deadline for registration is June 30, 2018.
Registration is now closed.
The training sessions will be held all day August 19 and 20.
Course on National Accounts
The course in based on the second edition (October 2014) of the OECD manual “Understanding National Accounts” by François Lequiller and Derek Blades. The manual is available free on line. The chapters referenced in the time-table are those of the manual.
The course is directed to non-experts. While it covers the theoretical definition of the main national accounts aggregates, it does so in a lively manner based on the concrete experience of the teacher (including as a number-cruncher), and on some exercises. François Lequiller has been the head of the French national accounts and has a wide international experience (OECD, IMF, Eurostat). He will explain the sources from which aggregates are derived, and, last but not least, will discuss their quality and their limitations. A part of the course will cover the themes that have been recently dubbed as “beyond GDP”: well-being and globalisation.
The duration of the course is of 16 hours, in 8 sessions.
Day 1 (August 19, 2018):
8:00-10:30 – Session 1: The essential macroeconomic aggregates: GDP, GNP, the basic accounting identities (Chapter 1)
11:00-12:30 – Session 2: The national accounts machinery, integrated economic accounts and the limitations of national accounts aggregates (Chapters 10 and 11)
13:30-15:30 – Session 3: The volume/price split in spatial and geographical comparisons, and international comparability (Chapters 2 and 3)
16:00-18:30 – Session 4: The production frontier and final uses (Chapters 4 and 5)
Day 2 (August 20, 2018):
8:00-10:30 – Session 5: Household and business accounts (Chapters 6 and 7)
11:00-12:30 – Session 6: Financial accounts (Chapter 8)
13:30-15:30 – Session 7: General government accounts (Chapter 9)
16:00-18:30 – Session 8: GDP, well-being and globalisation (Chapters 15 and 16)
The power points of the course will be circulated to the attendants one week before the course. A hand-held calculator would be useful.
Course on Advantages and Disadvantages of Non and Semi-parametric Methods for Data Classification and Categorization
Preamble
One major role of statistical agencies the world over is the analysis and classification of data. Generally, classification methods are based upon arbitrarily chosen (occasionally criticised) criteria. Non-parametric and semi-parametric classification methods afford agencies and researchers the ability to assess the extent, magnitude and importance of latent groupings within their data products thus facilitating examination of the relevance of existing categorization criteria. The course will be taught by professors Maria Grazia Pittau and Roberto Zelli from Sapienza University of Rome, Italy.
Aim
The aim of this class it to introduce the basic concepts of nonparametric and semi-parametric classification methods with particular emphasis on different types of income data (truncated, grouped data..). Real datasets will be used for the empirical analyses and codes in R will be provided.
Outline of the Training Course
1. Data Issues
2. Kernel density estimation
2.1 From histogram to kernel density
2.2 Density estimation with sampling weights
2.3 Bandwidth selection
2.4 Adaptive kernel estimation
2.5 Multivariate and conditional density
2.6 Kernel density estimation in R
3. Finite mixture models
3.1 Mixtures of distributions
3.2 Estimation procedure
3.3 Number of components and number of groups
3.4 Group profiles: covariates entering the mixing weights
3.5 Mixture of regressions: covariates entering the component distributions
3.6 Estimating mixtures in R