Creating a 2000 IES-LFS Database in Stata

PROVIDE Project Technical Paper 2005:1

The remainder of total transfer incomes and receipts is assumed to be domestic
transfers. The sum of variables hhtransinc and hhtransexp should in theory be the same,
since all domestic transfer receipts should exactly offset domestic transfer payments.⁴³
In reality reported transfer payments are lower than transfer receipts, and hence the
expenditure side has to be adjusted. The following equation is used:

. replace hhtransexp = hhtransexp + ((hhtransexp/sumhhtransexp)*sumhhtransnet)

After this adjustment the sum of net transfers (variable sumhhtransnet) is
recalculated and equals zero. Note that some of these adjustments are made only once
fixing.do has been run (see section 4.2.3). The final mean values of transfers to and
from the rest of the world and inter-household transfers are listed below.

The inter-household transfers sub-matrix reports on all net transfer flows between all
RHGs in the matrix. The IES 2000, however, does not supply any information on ‘to
whom’ and ‘from whom’ transfers were paid or received. The only information we have
is the total value of transfer receipts and payments of each household (or household
group). The mapping of transfers between specific households or household groups
cannot be done in Stata. Certain assumptions have to be made about how these
payments or receipts are distributed. In the appendix (section 7.1) we explain the
assumptions and show how the cells of the inter-household transfers matrix may be
populated using a simple MS Excel model.

4.2.3. Income and expenditure differences (fixing.do)

After running transfers.do, do-file fixing.do is run to close the gap between income and
expenditure. This gap is removed by assuming that the larger of income or expenditure
reported by each household is the correct welfare estimate. The under-declared figure is
increased while each of the components that make up the under-declared figure is
scaled upwards.

Do-file fixing.do starts with the income side and calculates each observation’s
undeclared income (variable incundecl) if income is less than expenditure and adds this
to incother. Total income is increased by incundecl. In order to ensure that the
components of income add up to the adjusted income figure, each component is scaled

⁴³ Because it is a ‘zero sum game’ national accounts ignore inter-household transfers. However, this data
is usually included in a SAM and hence it cannot be ignored here.

59

© PROVIDE Project

<<   <   61   62   63   64   65   66   67   >   >>