PROVIDE Project Technical Paper 2005:1
February 2005
4.2.1. Merging the IES 2000 and LFS 2000:2 files (ieslfsmerge.do)
Do-file ieslfsmerge.do uses the original versions of the person-level IES 2000
(ies2000p_orig.dta) and LFS 2000:2 (lfs2000_2_orig.dta) created in part one of
ies2000.do to form a merged file called ieslfsmerge.dta. Given some of the
discrepancies between the IES 2000 and LFS 2000:2 in terms of demographic variables,
the LFS 2000:2 variables are used. However, in some cases the demographic variables
are missing in the LFS 2000:2 but not in the IES 2000. In such cases the IES 2000
variables are used to replace missing LFS variables. This do-file creates variables for
factors (mergefact), labour income (mergeinclabp), activities (mergeact), gender
(mergegender), age (mergeage), province (mergeprov), location (mergeloc), race
(mergerace) and person-weights are created (mergepwgt).
4.2.2. Adjusting transfer variables (transfers.do)
Do-file transfers.do has two objectives. Firstly, it creates variables for transfers to and
from the rest of the world (rowtransinc and rowtransexp), and secondly, it addresses the
disparity between total transfer receipts and total transfer payments in IES 2000. The
initial mean values of transfer expenditure (hhtrans) and transfer receipts (inctrans) are
listed below.
Variable | Obs Weight Mean Std. Dev. Min Max
-------------+-----------------------------------------------------------------
hhtrans | 26265 11041643.4 851.5219 5480.98 0 360000
inctrans | 26265 11041643.4 1724.127 5838.838 0 396185
The IES 2000 does not distinguish between domestic and foreign transfer receipts or
payments. We make the assumption that the transfer receipts and payments reported are
inclusive of both domestic and foreign transfers. In order to separate out the foreign
transfers from total transfers, we make use of SARB 2000 data, which reports on total
foreign transfer receipts and payments to and from South Africa. These transfers make
up 0.04% and 0.02% of current household income respectively. This information is
used to estimate the share of foreign transfers in total transfer receipts and payments.
The distribution of foreign transfer payments or receipts is weighted according to each
household’s share of total (national) transfer receipts or payments, i.e. foreign transfers
follow the same distribution pattern as domestic transfers.42 The total sum of transfer
receipts is created so that it equals 0.04% of household income, while total transfer
payments is created so that it equals 0.02% of household income (national level).
42 This assumption is made due to the lack of information regarding the distribution of foreign transfers
across households. An alternative approach would be to assume that foreign transfer activities are
related to the income of the household, e.g. that a higher income household is more likely to
receive income or make payments to household abroad. However, since there is no real basis for
such an approach the former approach was opted for.
58
© PROVIDE Project