
Stata datafile (employments.dta, members_wms.dta, and WelfareWB.dta) obtained from Faiz'vi on 01 September 2010.

The working folder is c:\zunaid\bgd\mes2009\data. WelfareWB.dta in not included into the processed LM-MD dataset.

NOTE: merge command notes "variable indnum does not uniquely identify observations in the master data; variable indnum does not uniquely identify observations in c:\zunaid\bgd\mes2009\data\datawaste\EMP.dta" but could not find any duplicate individual id

Variables of the Health Module dropped.

For urban areas, variable RMO is matched with variable UR in the EMP dataset.

AGEY is checked with the created AGE_GROUP variable in EMP file.

Relationship variable Q3_3 has extra codes outside the listed ones.

This dataset doesn't have multiple head in same HH.

Household size variable already clreated - unusual values 45 and 101 checked and corrected.

Ethnicity info not collected in this survey.

Language info is not collected in this survey.

Religion variable collected in the intro section, not available in the datasets.

Info on marital status was collected from all the members in the household
. Married polygamy and living together info not collected; the "others" category needs to be checked. Others category has 31 females and 9 males.

Education level info collected for ages 5 years and above. Matching with age indicates that grade passed: PRIMARY group seems to be primary incomplete.
NOTE: edu level has an additional code (7), which falls outside the categories
NOTE: Observations of age <5 or coded as 7 are recoded as missing

Data not collected on completed years of education, only the education level completed - EDYEARS cannot be computed.

MES does not provide info for separating "Some Primary, but not completed" and "Completed Primary" groups - EDLEVEL_DAVID cannot be computed.

Not enough info for creating variable CONEDLEVEL and CONEDYEARS.

Q3_13 and LF1 collects information on employment, but doesn't match fully - LF1 has recorded a number of employed under 15 yrs. Q3.13  has 5 observations falling outside listed categories.

Reference period for Q3.13 is last 7 days.

Q3.13 did not collect information to construct discouraged category. It has two responses for Unemploed category - unemployed and looking for job.

A number of observations are under 15 years of age - recoded to missing.

In WHYINACTIVE, housewife recoded as "doing housework" and disabled recoded as "sick/incapable".

Q3_14 collects information with a created industry/occupation classification. As per the questionnaire, LF4 was supposed to contain 2 digit level codes as per BSIC 2001 (BSIC2001.pdf) but the variable has only values ranging from 0 to 5 (and 65, 93) and LF5 appears top contain industry info.

For SECTOR_MAIN, variable LF5 was used. 2-digit ISCO-88 codes (Meeting with Ainul Kabir on 04 Nov 2010) are used from ISCO.pdf.

BGD MES 2009 did not collect info on this NUMJOBS12MO.

For working hours, 0.10% and 0.01% respondents respectively stated the working hours for main and secondary jobs to exceed 96 per week.

Regional price deflator was created using 2007/08 CPI of urban and rural areas (file: CPI_January-09.pdf).

Info on individual income for main or secondary job is only available for day labour and paid worker. Original question LF11 (how do you get the salary/wages?) is not recorded.

Reference period is one week for INCOME_MAIN_def.

Outlier corrected for INCOME_MAIN_def, HHINCOME_TOT_def.

XINCOME_MAIN_def, XINCOME_TOT_def, XNONLBRINC_def, IMP_FM_RENT_def above variables are not relevant for BGD, as info on implicit land rental cost not collected.

WMS 2009 collected no consumption information. Hence TOTCONS_def, CONS_PC_def, etc. were not collected.

Poverty lines are calculated as average for urban and rural from HIES 2005 report (page 160, poverty lines.pdf)
.

ppp05deflator is set as 25.49 (source: Scoring_Poverty_Bangladesh_2005_EN.pdf).

Stratification done in two stages (p.13).

Weights are only available for working-age population.

No info collected on training. Hence EDLEVEL_VT, VT_CATEG were kept blank.

SCHOOL_LEAVE, CURRENT_ATTEND, DURATION_UNEMP, CASUAL_OR_WAGE, ENROL_CHILDREN, SCHOOL_DIST, PENSION, PENSION_INCOME, CONTRIBUTORY_HEALTH etc. are kept blank.
