Sonia Bhalotra and Cliff Attfield, "Intrahousehold Resource Allocation in Rural Pakistan: A Semiparametric Analysis", Journal of Applied Econometrics, Vol. 13, No. 5, 1998, pp. 463-480. There is a single data file, called ba-data.dat, which is zipped in the file ba-data.zip. The file is an ASCII file in DOS format. When unzipping it on a Unix system, use "unzip -a". Filesize: 4046901 (Unix format) 4183275 (DOS format) The file is organized as columns separated by spaces. The data are a cross section of households, and the columns are variables. The number of records is 136374. Min record length=7 and max=105. Number of observations=9741 Number of variables=700 Source of data: Household Income and Expenditure Survey, Government of Pakistan. LIST OF VARIABLES (slash denotes end of line in the data file): hhcode fdsh lsize lpcexp lpcexpsq/ lpcinc lpcincsq mksh adsh chsh/ dp dn ds db dq1 dq2 dq3 dq4/ dmorder dforder hhhlit hhslit/ prpm01-prpm08/ wprpm09 wprpm10-wprpm15 pmwork5-pmwork7/ dprpm09 dprpm10-dprpm15 pmdept5-pmdept7/ prpf01-prpf08/ wprpf09 wprpf10-wprpf15 pfwork5-pfwork7/ dprpf09 dprpf10-dprpf15 pfdept5-pfdept7/ prpm01-prpm09/ prpm10-prpm15 m5-m7/ prpf01-prpf09/ prpf10-prpf15 f5-f7. Variable definitions: hhcode = household ID fdsh = share of food expenditure in household budget lsize = log of household size lpcexp = log of per capita expenditure of household lpcexpsq = square of lpcexp lpcinc = log of per capita income of household lpcincsq = square of lpcinc mksh = share of milk expenditure in household budget chsh = share of budget spent on "child goods" adsh = share of budget spent on "adult goods" dp = dummy variable that is 1 if province=Punjab dn = dummy variable that is 1 if province=NWFP ds = dummy variable that is 1 if province=Sind dq1 = dummy variable that is 1 if quarter=1 dq2 = dummy variable that is 1 if quarter=1 dq3 = dummy variable that is 1 if quarter=1 dq4 = dummy variable that is 1 if quarter=1 dmorder = birth order dummy for males (dmorder is defined as 1 if there exists a male child in the hh) dforder = birth order dummy for females hhhlit = literacy (0/1) of household head hhslit = literacy (0/1) of household head's spouse prpm(i) = number of males of age i as a proportion of total hh size. prpf(i) = number of females of age i as a proportion of total hh size. wprpm(i) = number of working males of age i as propn. of hhsize wprpf(i) = number of working females of age i as propn of hhsize dprpm(i) = number of dependent males of age i as a propn of hhsize dprpf(i) = number of dependent females of age i as a propn of hhsize (NB: wprpm(i)+dprpm(i)=prpm(i), ie propn working males + propn dependent males = propn males.) pmwork(i) = number of working males of age-group i as a proportion of hhsize. pfwork(i) = number of working females of age-group i as a proportion of hhsize. pmdept(i) = number of dependent males of age-group i as a proportion of hhsize. pfdept(i) = number of dependent females of age-group i as a proportion of hhsize. Sonia R. Bhalotra Department of Economics University of Bristol 8 Woodland Road Bristol BS8 1TN U.K. E-mail: s.bhalotra@bristol.ac.uk Telephone: +44 117 928 8418 Fax: +44 117 928 8577