Creating a 2000 IES-LFS Database in Stata

PROVIDE Project Technical Paper 2005:1
(standard 8), (4) upper secondary (standard 10), (5) tertiary and (6) other, don’t know or
missing. These education categories are used later in the formation of representative
household groups for the SAM (see PROVIDE, 2005).

Sub-do-file factact.do creates a wage or salary-income variable (w_inclabp), an
occupation code variable (w_fact) and an activity variable (w_activity). Respondents had the
choice of either specifying their exact weekly, monthly or annual wage or salary, or selecting
an income category within a specific income band (discrete variable). Consequently it was
necessary to adjust the data so that variable w_inclabp is a continuous variable showing the
respondent’s annual income from labour. Variable w_fact has the same occupation categories
as listed in Table 7, while variable w_activity is similar to variable activities created in the
IES 2000.

4. Further data analysis and adjustments

The second part of ies2000.do starts by running adjustments.do. This do-file can also be run
independently from ies2000.do. Do-file adjustments.do adjusts the data in ies2000h.dta and
ies2000p.dta (or ieslfsmerge.dta, a merged IES 2000 and LFS 2000:2 dataset) so that they are
ready to use for the creation of various sub-matrices in a series of SAMs for South Africa.
These SAMs are compiled from a variety of datasets, using sources of national data, such as
the SARB data, to provide control totals. This is necessary so that the SAM represents a
realistic picture, not only of patterns of income and expenditure of agents, but also of the
overall income and expenditure levels.

The main objective of adjustments.do is to balance incomes and expenditure of individual
households. By definition income and expenditure of each household group in the SAM
should balance. Although the final balancing of a SAM is done during the SAM estimation
process (PROVIDE, ____-a), balancing incomes and expenditures at an individual

household-level at this stage in the process if also useful as it ensures that households either
income or expenditure can be used to determine the household’s level of welfare.³⁹ Before
proceeding with the discussion adjustments.do, section 4.1 looks at the income and
expenditure differences at the household level.

4.1. Income and expenditure differences

In theory total income and expenditure reported in the IES 2000 should be the same due to the
way the questionnaire is set up and because of the economic identity
income = consumption + savings. However, this is not true for the IES 2000. This can be due

³⁹ For example, if household groups were formed on the basis of income or expenditure, large differences
between income and expenditure would cause households to be allocated to very different groups.

52
© PROVIDE Project

<<   54   55   56   57   58   59   60   >>

More intriguing information