DASP Distributive Analysis Stata Package By Abdelkrim Araar

1
DASPDistributive Analysis Stata
PackageByAbdelkrim Araar
2
Main objective of the DASP project

• Main objective
• The main aim of this project is to produce a
  complete package of STATA modules to analyze the
  distribution of living standards. This is
  important for measurement as well as policy
  purposes.
•  
• Why use the STATA software?
• The STATA software has become in the last 20
  years a very popular tool to transform and
  process data. It comes with a large number of
  basic data management modules that are highly
  efficient for transformation of large datasets.
  The flexibility of STATA also enables programmers
  to provide specialised .ado routines to add to
  the power of the software.

3
Why DASP is very helpful?

• DASP (Distributive Analysis STATA Package) can be
  used easily by all researchers that are
  interested to produce results and to perform the
  distributive analysis.
• Most interest, this package should be very
  helpful for researchers in developed countries
  that can facilitate computation   of the desired
  distributive indices, as well as, the application
  of the most recent approaches in this field by
  maintaining an updated form of this package.

4
What is already available in STATA for
distributive analysis?

• Some STATA .ado files already exist for the
  computation of some specific distributive indices
  or for plotting some distributive curves.
• Furthermore, the few available modules has the
  following inconvenient 
• They do not have an unified syntax
• In some cases, basic option like to weight the
  data are not available
• Standard errors are not furnished or do not take
  into account the design effect.  
• Programs are not optimised to run quickly and
  some bug can be encountered.

5
In which way does DASP differ from these
available modules?

• The modules of DASP would be designed to
• Cover the most popular indices and curves in the
  field of distributive analysis.
• Support the analysis by using more than one data
  base
• Perform the most popular decomposition of poverty
  or inequality indices
• Unify the syntax and the provision of parameters.
  
• Other softwares and DASP
• DAD
• Free.
• Very user friendly for the user
• Has a broad coverage of distributive analysis.
• A limitation is the maximal number of variables
  (20)
• It is not designed to provide basic data
  processing tools.
• Does not support missing values.

6
In which way does DASP differ from these
available modules?

• The modules of DASP would be designed to
• Cover the most popular indices and curves in the
  field of distributive analysis.
• Support the analysis by using more than one data
  base
• Perform the most popular decomposition of poverty
  or inequality indices
• Unify the syntax and the provision of parameters.
  
• Other softwares and DASP
• DAD
• Free.
• Very user friendly for the user
• Has a broad coverage of distributive analysis.
• A limitation is the maximal number of variables
  (20)
• It is not designed to provide basic data
  processing tools.
• Does not support missing values.

7
In which way does DASP differ from these
available modules?

• POVCAL
• Covers few tools and indices
• Is not particularly user friendly
• Is not designed to treat micro data.

8
DASP PEP Projects

• Many PEP projects, for the wave 2005/06, have
  already beneficed from the DASP package.
• The advantage of using directly these modules
  with Stata are
• Exploiting all facilities of Stata for the
  treatment of the data.
• Writing do programs to keep all procedures of the
  treatment of the data, variables transformations,
  recoding, etc.
• Writing the do file for all estimations
• This organisation allow to the researchers to
  improve or correct theirs estimations with few
  time and energy

9
Recommanded Organisation
10
Already realised in DASP

• Distributive Indices
• ifgt.ado
• isgini.ado
• Distributive curves
• clorenz.ado
• clorenz can produces the following distributional
  curves for a given list of variables
• Lorenz curves
• Generalised Lorenz curves
• Absolute Lorenz curves
• Concentration curves
• Generalised concentartion curves
• Absolute Concentration curves
• Deficit share curves (p-L(p))
• Diffrence between Lorenz curves

11
Already realised in DASP

• Distributive curves
• cfgt.ado
• can produces the following distributional curves
  for a given list of variables
• FGT curves (Foster, J. E., J. Greer , and E.
  Thorbecke curves)
• Normalised FGT curves
• cquantile.ado
• can produces the following distributional curves
  for a given list of variables
• Quantile curves
• Normalised quantile curves
• Absolute quantile curves
• cdepriv.ado
• can produces the following distributional curves
  for a given list of variables
• Deprivation curves
• Absolute Deprivation curves

12
Already realised in DASP

• Distributive curves
• cdensity.ado
• Basing on the Kernel-Gaussian approach, cdensity
  produces the density curves for a given list of
  variables or according to population groups.
• cnpe.ado
• produces the non parametric regression curves for
  a given list of variables or according to
  population groups according to one of the two
  following approaches
• Nadaraya-Watson approach is used. One can also
  perform regressions with the
• Linear locally approach

13
Already realised in DASP

• Poverty dominance
• povdom.ado
• Check the poverty dominance for the first, second
  and the third order of dominance in poverty.
  Estimates all possible intersections
• Inequality dominance
• ineqdom.ado
• Check the inequality dominance basing on the
  comparison between the Lorenz curves. Estimates
  all possible intersections
• Consumtion dominance curves
• ccdom.ado
• ccdom produces the consumption dominance curves
  (CD) and normalised CD curves for a given list of
  variables or according to population groups.

14
Already realised in DASP

• Polarisation
• Duclos, Esteban Ray Index (poder.ado)
• Decomposition
• Transient and chronic poverty
• Jalan Ravallion approach (dtcjr.ado)
• EDE approach (dtcda.ado)