******************************************************************************************************** * Syntax file to accompany 'Work, Attitudes & Spending' household surveys: * * No need to run this file again (you can read it, to see how variables are derived). * ********************************************************************************************************. * Default value labels: use these for all numeric variables. They are used in BHPS, except -6. *-1"don't know" -2"refused" -6"not asked in this survey" -8"inapplicable" -9"missing/wild [e.g. no answer]". *-7"proxy respondent": I never use it. -3 & -4 specific to a particular variable, e.g. too young for school. *@ INTERVIEW DETAILS. * Create a unique ID number for each household. Ignore ID: it's just temporary. if (INT_year= 1992) surveyN = 1. if (INT_year= 1994) surveyN = 2. if (INT_year= 1997) surveyN = 3. if (INT_year= 2000) surveyN = 4. if (INT_year= 2001) surveyN = 5. if (INT_year= 2002 & live_cit> 6) surveyN = 6. if (INT_year= 2002 & live_cit<10) surveyN = 7. if (INT_year= 2003) surveyN = 8. if (INT_year= 2004) surveyN = 9. if (INT_year= 2005) surveyN = 10. if (survey ='2005-6 Egypt') surveyN = 11. if (INT_year= 2007) surveyN = 12. if (INT_year= 2008) surveyN = 13. if (INT_year= 2009) surveyN = 14. if (INT_year= 2011) surveyN = 15. if (INT_year= 2012) surveyN = 16. if (INT_year= 2017) surveyN = 17. var label surveyN 'Survey number'. compute ID = 1. /* Create ID, to ensure consecutive numbers. exec. if (surveyN = lag(surveyN)) ID = lag(ID) + 1. compute IDnum= (surveyN*100000) + ID. var label IDnum 'Household identification number'. var lab temperat 'Temperature at the time & place of interview (measured by thermometer)'. var lab thermometer 'Type of thermometer used, to measure the temperature'. val lab thermometer 0 'Weather genius' 1 'Magna-temp'. /* 'Weather genius' has a digital display with 1 decimal place; 'Magna-temp' only whole numbers. val label INT_mnth 1"January" 2"February" 3"March" 4"April" 5"May" 6"June" 7"July" 8"August" 9"September" 10"October" 11"November" 12"December". var label INT_year 'Year of interview'. var label INT_mnth 'Month of interview'. var label INT_day 'Day of month of interview'. var label INT_time 'Time of day (nearest hour) the interview ended'. var label INTplace 'Place where interview took place'. val label INTplace 1 'inside the home' 2 'on the doorstep' 3 'somewhere else'. var label INT_CHEK 'Was interview checked by a (market research organisation) supervisor?'. val label INT_CHEK 0"not checked" 1"accompanied by supervisor" 2"backchecked" 3"saw form" 4"accompanied & backchecked" 5"accompanied & saw form" 6"saw form & backchecked" 7"accompanied, saw form & backchecked". * First, a simplified version of INT_PRES (INT_PRES not recorded in Indonesia, but PRESENT was recorded). if (INT_PRES>1) INTpresn = 1. if (INT_PRES=1) INTpresn = 0. var label INTpresn 'Was anyone (exept respondent) present at interview?'. val label INTpresn 0'respondent only' 1'1 or more others present'. var label INT_PRES 'If anyone except respondent present at interview, who?'. val label INT_PRES 1'respondent only' 2'spouse' 3'other adult(s)' 4'child(ren)' 5'spouse and other' 6'spouse & child(ren)' 7'children/other adult' 8'spouse/child/other'. var label INT_CODE 'Code for interviewer'. var label INTfemal 'Gender of interviewer'. val label INTfemal 0 'male interviewer' 1'female interviewer' 0.5'male & female interviewers'. var label SEX 'Gender of respondent'. val label SEX 0'male' 1'female'. * p1_relat etc won't report as HoH if p1 is respondent. value labels fem_HEAD 0'male' 1'female'. var label fem_HEAD "Female headed household". var lab resp_ORD 'Respondent mentioned himself/herself 1st/2nd/etc'. var lab spou_ORD 'Respondent mentioned his/her spouse 1st/2nd/etc'. var lab CWE_ORD 'Respondent mentioned Chief Wage Earner 1st/2nd/etc'. val label resp_ORD 1"mentioned self 1st" 2"mentioned self 2nd" 3"mentioned self 3rd" 4"mentioned self 4th" 5"mentioned self 5th" 6"mentioned self 6th" 7"mentioned self 7th" 8"mentioned self 8th" 9"mentioned self 9th" 10"mentioned self 10th" 11"mentioned self 11th" 12"mentioned self 12th" 13"mentioned self 13th" 14"mentioned self 14th" 15"mentioned self 15th" 16"mentioned self 16th" 17"mentioned self 17th". val label spou_ORD 1"mentioned spouse 1st" 2"mentioned spouse 2nd" 3"mentioned spouse 3rd" 4"mentioned spouse 4th" 5"mentioned spouse 5th" 6"mentioned spouse 6th" 7"mentioned spouse 7th" 8"mentioned spouse 8th" 9"mentioned spouse 9th" 10"mentioned spouse 10th" 11"mentioned spouse 11th" 12"mentioned spouse 12th" 13"mentioned spouse 13th" 14"mentioned spouse 14th" 15"mentioned spouse 15th". val label CWE_ORD 1"mentioned CWE 1st" 2"mentioned CWE 2nd" 3"mentioned CWE 3rd" 4"mentioned CWE 4th" 5"mentioned CWE 5th" 6"mentioned CWE 6th" 7"mentioned CWE 7th" 8"mentioned CWE 8th" 9"mentioned CWE 9th" 10"mentioned CWE 10th" 11"mentioned CWE 11th" 12"mentioned CWE 12th" 13"mentioned CWE 13th" 14"mentioned CWE 14th" 15"mentioned CWE 15th". *cros p1_relat by survey. /* To identify which surveys we can reliably for wifBefor & husBefor. *if (survey='1992 India') resp_ORD = -9. /* The respondent is USUALLY first in 1992 India; so wifBefor & husBefor unreliable. *if (survey='1992 India') spou_ORD = -9. *if (survey='1994 Brazil') resp_ORD = -9. /* The respondent is USUALLY first in 1994 Brazil; so wifBefor & husBefor unreliable. *if (survey='1994 Brazil') spou_ORD = -9. *if (survey='1997 India') resp_ORD = -9. /* The respondent is USUALLY first in 1997 India; so wifBefor & husBefor unreliable. *if (survey='1997 India') spou_ORD = -9. *if (survey='2002 India') resp_ORD = -9. /* The respondent is USUALLY first in 2002 India; so wifBefor & husBefor unreliable. *if (survey='2002 India') spou_ORD = -9. *if (survey='2003 Nigeria') resp_ORD = -9. /* The respondent is USUALLY first in 2003 Nigeria; so wifBefor & husBefor unreliable. *if (survey='2003 Nigeria') spou_ORD = -9. if (survey='2004 Kenya') resp_ORD = -9. /* The respondent is ALWAYS first in 2004 Kenya; so wifBefor & husBefor unhelpful. if (survey='2004 Kenya') spou_ORD = -9. *if (survey='2005 Nigeria') resp_ORD = -9. /* The respondent is USUALLY first in 2005 Nigeria; so wifBefor & husBefor unreliable. *if (survey='2005 Nigeria') spou_ORD = -9. if (married=0 or (married>2 & married<7)) HUSbefor = -4. if (married=0 or (married>2 & married<7)) WIFbefor = -4. recode resp_ORD spou_ORD ( -9 = sysmis). if (sex=1 & resp_ORD>spou_ORD) HUSbefor = 1. if (sex=1 & resp_ORDspou_ORD) WIFbefor = 1. if (sex=0 & resp_ORD40) EDUC = -9. /* ...implausible, so could set EDUC to missing; but maybe literate is incorrect?. recode educ_hus educ_wif (sysmis = -77). /* educ_hus & educ_wif are known for India 1992 (at least). if (educ_wif = -77 & sex=0) educ_wif = EDUCspou. if (educ_wif = -77 & sex=1) educ_wif = EDUC. if (educ_hus = -77 & sex=0) educ_hus = EDUC. if (educ_hus = -77 & sex=1) educ_hus = EDUCspou. recode educ_hus educ_wif (-77 = sysmis). var label EDUC "Education level of respondent". var label EDUCspou "Education level of spouse of respondent". var label educ_hus "Education level of husband of respondent". var label educ_wif "Education level of wife of respondent". var label EDUC_CWE "Education level of Chief Wage Earner". val label EDUC, EDUCspou, educ_hus, educ_wif, EDUC_cwe 0 "no school/illiterate" 10 "school up to 4 years" 20 "over 4 years primary school" 30 "primary school completed, some secondary" 40 "secondary school, incomplete/preparatory school" 55 "matric/Secondary School Cert/O or A-level" 57 "artisan's certificate" 65 "college not graduate (eg technical)" 72 "college: diploma/secretarial qual" 75 "higher college (Technikon) diploma/degree" 77 "incomplete degree" 80 "graduate/above, general" 90 "postgrad (e.g. MA); graduate professional" 100 "PhD" 59 "other" -2 "not disclosed" -3 "too young to be at school". /* My scale: 0=no school to 100=20 years (i.e. PhD). * Educ=30 (formerly =35) should be equivalent to '5 to 9 years of educ' in IMRB 1997 & 2002. * In Indonesia,'college diploma' means 'college diploma D1/D2'; 'higher college diploma' is 'college diploma D3'. * For Indonesia, rs09 indicates if study completed - could use this to make more detailed educ variable. * Make WAS comparable to DHS, in having a variable to estimate the number of years in education. I drop 59 (although this reduces the sample-size). if (LIVEctry<>'CM') educY = educ. /* For Cameroon 2009 survey, respondents reported (approx) exact number of years in education. if (LIVEctry<>'CM') educYhus = educ_hus. if (LIVEctry<>'CM') educYwif = educ_wif. var label educY "Number of years in education: respondent". var label educYhus "Number of years in education: husband /male respondent". var label educYwif "Number of years in education: wife /female respondent". recode educYwif educYhus educY (0=0)(10=3)(20=5)(30=7)(35=7)(40=8)(55=11)(65=12)(72=13)(75=14)(77=15)(80=16)(90=17)(100=20)(59=sysmis). * The above recodes are arbitrary, and are based on Cameroon. * Ideally, should go back to each WAS survey and change the above line to an appropriate version for that questionnaire. if (LIVEctry = 'CM' & sex=0) educYhus = EDUCy. if (LIVEctry = 'CM' & sex=1) educYwif = EDUCy. if (LIVEctry = 'CM' & sex=1) educYhus = EDUCspouy. if (LIVEctry = 'CM' & sex=0) educYwif = EDUCspouy. recode EDUC, educY, educYwif, educYhus ( -77, 98 =sysmis). missing values EDUC, educY, educYwif, educYhus ( -9 thru -1). * Languages: I don't put the next block of 30 lines in 'language and ethnic vars.sps', because that files is also used for DHS and Afrobarometer. var label L_Englis 'Know language: English'. var label L_Afrika 'Know language: Afrikaans'. /* first of South African languages. var label L_Zulu 'Know language: Zulu'. var label L_Xhosa 'Know language: Xhosa'. var label L_NSotho 'Know language: N. Sotho'. var label L_SSotho 'Know language: S. Sotho'. var label L_Tswana 'Know language: Tswana'. var label L_Tsonga 'Know language: Tsonga/Shangaan'. var label L_Venda 'Know language: Venda'. var label L_Swazi 'Know language: Swazi'. var label L_Ndebel 'Know language: Ndebele'. /* last of South African languages. var label L_Hausa 'Know language: Hausa'. /* first of Nigerian languages. var label L_Igbo 'Know language: Igbo'. var label L_Yoruba 'Know language: Yoruba'. var label L_Fulani 'Know language: Fulani'. var label L_Edo 'Know language: Edo'. var label L_Tiv 'Know language: Tiv'. var label L_Nupe 'Know language: Nupe'. var label L_Urhobo 'Know language: Urhobo'. /* last of Nigerian languages. var label L_other 'Know language: other(s)'. val lab L_Englis L_Afrika L_Zulu L_Xhosa L_NSotho L_SSotho L_Tswana L_Tsonga L_Venda L_Swazi L_Ndebel L_Hausa L_Igbo L_Yoruba L_Fulani L_Edo L_Tiv L_Nupe L_Urhobo L_other 0"can't read this langauge" 1"can read this language" 2"understand it, but can't read it" 3"can both read & understand it". *freq L_Englis. * These value labels are already in the WAS surveys which have literate (ZA 2000, NG 2003, and NG 2005). *var label literate "Respondent's literacy". /* DHS use all labels below; WAS only uses 0 or 2. *val label literate 0 'cannot read' 1'able to read parts of sentence' 2 'able to read whole sentence' -3 'no card with required language' -4 'blind/visually impaired' -9 "missing data". *missing values literate (-9 thru -1). *@ WORK / SOCIAL CLASS. *EDUC_CWE and JOB_CWE combined into one summary measure, by IMRB (CWE_CODE): . *var label CWE_CODE 'IMRB classification of Chief Wage Earner'. *cros CWE_code by survey. /* zzz these next lines may be already run, in each survey. if (CWE_code = 'A1') IMRBclas = 1. if (CWE_code = 'A2') IMRBclas = 2. if (CWE_code = 'B1') IMRBclas = 3. if (CWE_code = 'B2') IMRBclas = 4. if (CWE_code = 'C ') IMRBclas = 5. if (CWE_code = 'D ') IMRBclas = 6. if (CWE_code = 'E1') IMRBclas = 7. if (CWE_code = 'E2') IMRBclas = 8. *val lab IMRBclas 1 'A1' 2 'A2' 3 'B1' 4 'B2' 5 'C' 6 'D' 7 'E1' 8 'E2'. val lab IMRBclas 1 'A1' 2 'A2' 3 'B1' 4 'B2' 5 'C' 6 'D' 7 'E1' 8 'E2' 1.5 'AB' 5.1 'C1' 5.2 'C2'. * The last 3 values are for Cameroon; this is Cible, not IMRB. var lab IMRBclas "Class of household, based on head-of-household's job & education (IMRB and Cible)". *cros IMRBclas by survey. * Details, from IMRB India 2002-2007 questionnaires: EDUCATION * OCCUPATION ILLITERATE SCHOOL SCHOOL SSC/ COLLEGE GRAD/ GRAD/ * UP TO 4 YRS 5-9 YRS HSC NOT GRAD POSTGRAD POSTGRAD * (GEN) (PROF) * 1 2 3 4 5 6 7 * ---------------------------------------------------------------------------------------------------------------------------------------- * Unskilled E2 E2 E1 D D D D * Skilled worker E2 E1 D C C B2 B2 * Petty trader E2 D D C C B2 B2 * Shop owner D D C B2 B1 A2 A2 * Businessmen: * no employees D C B2 B1 A2 A2 A1 * * 1-9 employees C B2 B2 B1 A2 A1 A1 * *10+ employees B1 B1 A2 A2 A1 A1 A1 * self-employed professional D D D B2 B1 A2 A1 * clerical/sales D D D C B2 B1 B1 * supervisory level D D C C B2 B1 A2 * officer/exec: junior C C C B2 B1 A2 A2 * officer/exec: middle/senior B1 B1 B1 B1 A2 A1 A1 if (sex=0) job_hus = job. if (sex=1) job_hus = job_spou. if (sex=1) job_wif = job. if (sex=0) job_wif = job_spou. var label job "Job code of respondent". var label job_spou "Job code of spouse of respondent". var label job_wif "Job code of resp's wife, or respondent if female". var label job_hus "Job code of husband of resp, or resp if male". var label job_CWE "Job of Chief Wage Earner". val label job_CWE, job, job_spou, job_hus, job_wif 98 "mid/senior officer/executive", 95 "officer/executive", 92 "junior officer/executive" 86 "manager: 10+ employees", 84 "manager:1-9 employees" 83 "manager", 82 "manager: 0 employees", 75 "professional (often self-employed)", 65 "self-employed: unspecified", 59 "shop owner", 51 "petty trader/hawker", 45 "supervisory", 40 "civil servant", 35 "clerical/sales", 30 "traditional doctor", 29 "manual: skilled", 28 "farmer", 27 "employed: unspecified", 25 "manual: unspecified", 24 "manual: semi-skilled", 21 "manual: unskilled", 20 "farm worker", 18 'clergy', 15 "student", 9 "housewife: was employed", 8 "housewife: dk if ever employed", 7 "housewife: never employed", 5 "unemployed: was employed", 4 "unemployed: never employed", 3 "unemployed: no details", 2 "unemployed: seeking work", 1 "unemployed: not seeking work", 0 "retired", -3 "too young to be employed", -2 "not disclosed", -1"don't know", -9 "missing/no answer/wild". * NOTE: The above value labels are also used for p1_job, p2_job, etc ('derive HH vars from P1 etc.sps'). * For Indonesia 2001 & 2002, associate professionals included in 45; maybe should be a new category. recode job job_resp (sysmis = -77). *cros job by job_resp. /* Cameroon only; this is a fairly good match, so delete variable job_resp. if (job= -77) job = job_resp. recode job job_resp ( -77 = sysmis). var label live_cit 'City where respondent lives'. *crosstab paid_hus paid_wif by live_cit. * WARNING: in Patna 2002, the hours of paid work is small (often 0) for both husband & wife. This isn't a data processing error: they're written (very clearly) on paper. * Maybe the question was mis-translated in Patna. My guess: maybe work hours are per day (not per week) in Patna? I tried this, but husband hours still low. * Maybe days per week in Patna? (most, but not all, cases are under 8). If so, how many hours per day: is 7 about right?. *if ( paid_hus>0) test_hus = paid_hus. *if ( paid_wif >0) test_wif = paid_wif. *if (live_cit=5 & paid_hus>0) test_hus = paid_hus*7. /*Assume people work 7 days per week,or 7 hours per day. *if (live_cit=5 & paid_wif >0) test_wif = paid_wif *7. *breakdown test_hus,test_wif by live_cit. * The above code seems to almost work. *if (live_cit=5 & paid_hus>0) paid_hus = paid_hus*7. *if (live_cit=5 & paid_wif >0) paid_wif = paid_wif *7. * But do another check: . *cros paid_hus by live_cit by job_hus. /*(job_hus calculated in 'derive HH vars from p1 etc.sps' & 'derive_v.sps'): many in Patna report paid_hus=0, * when they do have a paid job - so paid_hus data is not safe to use. if (live_cit=5) paid_hus = -6. /* Set to missing (seems a waste of data: e.g. zero hours=unemployed). if (live_cit=5) paid_wif = -6. *crosstab paid_hus paid_wif by live_cit. /*The next few lines set paid_hus & paid_wif to zero or missing. * In Nigeria (at least), there'e a lot of missing earnings data, and few earn zero: clarify using job_hus, job_wif. *breakdown earn_hus,paid_hus, earn_wif,paid_wif by workst. * PAID_hus, PAID_wif have far less missing data, than EARN_hus & EARN_wif. * Assume husb has no job if PAID_hus= 0 & EARN_hus missing. * Assume wife has no job if PAID_wif = 0 & EARN_wif missing. recode EARN_hus, EARN_wif ( sysmis= -90). if (EARN_hus<0 & PAID_hus = 0) EARN_hus = 0. if (EARN_wif <0 & PAID_wif = 0) EARN_wif = 0. if (EARN_hus<0 & job_hus > -1 & job_hus< 21) EARN_hus = 0. if (EARN_wif <0 & job_wif > -1 & job_wif < 21) EARN_wif = 0. recode EARN_hus, EARN_wif ( -90 = sysmis). if (survey='1992 India') EARN_hus = -99. /* All valid cases are zero; exclude them. if (survey='1992 India') EARN_wif = -99. /* All valid cases are zero; exclude them. if (survey='2000 S. Africa' & LIVEurbn>1) EARN_hus = -99. /* All valid cases are zero; exclude them. if (survey='2000 S. Africa' & LIVEurbn>1) EARN_wif = -99. /* All valid cases are zero; exclude them. recode EARN_hus EARN_wif ( -99 = sysmis). breakd EARN_hus EARN_wif by survey /cells mean. * Now, deal with missing time-use data. if (PAID_hus< 0 & job_hus > -1 & job_hus < 21) PAID_hus = 0. if (PAID_wif < 0 & job_wif > -1 & job_wif < 21) PAID_wif = 0. * For Indonesia, don't change the data (resp & spouse info available separately; I assume this is more reliable). *if ( (live_cit<41 or live_cit>44) & sex=0) paidResp = paid_hus. *if ( (live_cit<41 or live_cit>44) & sex=1) paidResp = paid_wif. *if ( (live_cit<41 or live_cit>44) & sex=1 & married=1) paidSpou = paid_hus. *if ( (live_cit<41 or live_cit>44) & sex=0 & married=1) paidSpou = paid_wif. * Above 4 lines are in 'derive data about the respondent.SPS', for all surveys. What about [married=7]?. * In future WAS surveys: make sure all time-use variables (cooking, etc) hours per week, not minutes per day. var label PAID_HUS "Hours of paid work (per week): self/husband". var label PAID_WIF "Hours of paid work (per week): self/wife". var label sEmp_HUS "Hours of self-employment (per week): self/husband". var label sEmp_WIF "Hours of self-employment (per week): self/wife". * Note: don't keep {washResp,washSpou}: just keep {wash_HUS, wash_WIF} (& other time-use vars). * for consistency, don't keep { paidResp, paidSpou}: just keep { paid_HUS, paid_WIF}. * If users want to find paidResp etc, they can run 'derive data about the respondent.SPS'. * For sa2000, WORKST includes information on FT vs PT job - very different to paid_hus/paid_wif. * I drop workst; it could be kept - but no equivalent for other years. Unclear what 'full-time' means (subjective). *if (isfemale=0) hours=paid_hus. *if (isfemale=1) hours=paid_wif. *crosstab hours by workst. missing value WHYnoJOB (' '). var label WHYnoJOB '[IF NOT EMPLOYED] Why are you not employed?'. val label WHYnoJOB '1' "baby/child" '2' "other unspecified" '3' "after marriage" '4' "family won't allow" '5' "family commitment/no time" '6' "pay would be low" '7' "prefer to mind kids" '8' "ill/retired/too old" '9' "can't get a job" '0' "no need to work" 'A' "husband won't allow" 'B' "no maid/can't get maid" 'C' "housewife" 'D' "student" 'E' "hoping to buy a business" 'F' "religion(women born to mind kids)" 'G' "minding sick mother". if (CLASS_WK >-1) CLASS = 0. if (CLASS_WK = 1) CLASS = 1. if (CLASS_SE = 1) CLASS = 2. if (CLASS_MI = 1) CLASS = 3. var label CLASS 'Highest social class of all HH members'. val label CLASS 0"student/not employed",1"working-class", 2"self-employed trader", 3"middle-class". var label CLASS_WK 'Is any household member working-class?'. val label CLASS_WK 0"not working-class" 1"working-class member". var label CLASS_MI 'Is any household member middle-class?'. val label CLASS_MI 0"not middle-class" 1"middle-class member". var label CLASS_SE 'Is any household member self-employed?'. val label CLASS_SE 0"no trader in hhold" 1"s-emp trader in hhold". *@ TIME-USE. * COOKING is split into 2 vars are in Brazil & Indonesia, but not in other surveys. if (COOKpHUS >-1 & COOKcHUS>-1) COOK_HUS = COOKpHUS + COOKcHUS. if (COOKpWIF >-1 & COOKcWIF>-1) COOK_WIF = COOKpWIF + COOKcWIF. *rs94 (only) had 2 cooking questions: COOKp & COOKc. COOK_HUS = COOKpHUS + COOKcHUS approx. missing values choreHus laundHus cook_hus wash_hus shop_hus mind_hus (-9 thru -1). missing values choreWif laundWif cook_wif wash_wif shop_wif mind_wif (-9 thru -1). *if (INT_year >1992) wash_HUS = laundHUS + cleanHUS. /* For comparability with bm92. *if (INT_year >1992) wash_WIF = laundWIF + cleanWIF. /* For comparability with bm92. if (INT_year >1992 & LIVEctry<>'ID') wash_HUS = laundHUS + cleanHUS. /* For comparability with bm92. if (INT_year >1992 & LIVEctry<>'ID') wash_WIF = laundWIF + cleanWIF. /* For comparability with bm92. breakd cook_hus wash_hus clothhus shop_hus mind_hus by survey/cells mean. compute hwrk_hus = cook_hus + wash_hus + shop_hus + mind_hus. /* For most surveys: exceptions listed below. if (survey='1992 India') hwrk_hus = cook_hus + wash_hus. /* shop_hus & mind_hus aren't in bm92; could remove this line. if (survey='1997 India') hwrk_hus = cook_hus + wash_hus + clothhus + shop_hus+ mind_hus. /* clothhus: India only. if (survey='2002 India') hwrk_hus = cook_hus + wash_hus + clothhus + shop_hus+ mind_hus. /* clothhus: India only. if (survey='2007 India') hwrk_hus = cook_hus + wash_hus + clothhus + shop_hus+ mind_hus. /* clothhus: India only. if (survey='2012 India') hwrk_hus = cook_hus + wash_hus + clothhus + shop_hus+ mind_hus. /* clothhus: India only. if (survey='2017 India') hwrk_hus = cook_hus + wash_hus + clothhus + shop_hus+ mind_hus. /* clothhus: India only. *if(survey='2002 Indonesia') hwrk_hus = cook_hus + wash_hus + shop_hus. /* ID 2002: mind_hus unavailable (data-processing problem). if (survey='1994 Brazil') hwrk_hus = cook_hus + laundhus + shop_hus+ mind_hus. /* BR 1994: included laundHus but not cleanHus. if (survey='2005-6 Egypt') hwrk_hus = choreHus + shop_hus+ mind_hus. /* choreHus includes various types of housework. *. *breakd cook_wif wash_wif clothwif shop_wif mind_wif by survey /cells mean. compute hwrk_wif = cook_wif + wash_wif + shop_wif + mind_wif. /* For most surveys: exceptions listed below. if (survey='1992 India') hwrk_wif = cook_wif + wash_wif. /* shop_wif & mind_wif aren't in bm92; could remove this line. if (survey='1997 India') hwrk_wif = cook_wif + wash_wif + clothwif + shop_wif + mind_wif. /* clothwif: India only. if (survey='2002 India') hwrk_wif = cook_wif + wash_wif + clothwif + shop_wif + mind_wif. /* clothwif: India only. if (survey='2007 India') hwrk_wif = cook_wif + wash_wif + clothwif + shop_wif + mind_wif. /* clothwif: India only. if (survey='2012 India') hwrk_wif = cook_wif + wash_wif + clothwif + shop_wif + mind_wif. /* clothwif: India only. if (survey='2017 India') hwrk_wif = cook_wif + wash_wif + clothwif + shop_wif + mind_wif. /* clothwif: India only. *if(survey='2002 Indonesia') hwrk_wif = cook_wif + wash_wif + shop_wif. /* ID 2002: mind_wif unavailable (data-processing problem). if (survey='1994 Brazil') hwrk_wif = cook_wif + laundwif + shop_wif + mind_wif. /* BR 1994: included laundwif but not cleanwif. if (survey='2005-6 Egypt') hwrk_wif = choreWif + shop_wif + mind_wif. /* choreWif includes various types of housework. * Housework time in India 1992 is much lower than later India surveys, because fewer time-use questions asked in 1992; could make hwkr_hus&hwrk_wif missing in 1992. * It's not clear if hwrk should include shopping or childcare. Calculate another variable, which excludes both; call it choreHus & choreWif (already in WAS EG). if (survey<>'2005-6 Egypt') chorehus = cook_hus + wash_hus. /* For most surveys: exceptions listed below. if (survey='1997 India') chorehus = cook_hus + wash_hus + clothhus. /* clothhus: India only. if (survey='2002 India') chorehus = cook_hus + wash_hus + clothhus. /* clothhus: India only. if (survey='2007 India') chorehus = cook_hus + wash_hus + clothhus. /* clothhus: India only. if (survey='2012 India') chorehus = cook_hus + wash_hus + clothhus. /* clothhus: India only. if (survey='2017 India') chorehus = cook_hus + wash_hus + clothhus. /* clothhus: India only. if (survey='1994 Brazil') chorehus = cook_hus + laundhus. /* BR 1994: included laundHus but not cleanHus. *. if (survey<>'2005-6 Egypt') chorewif = cook_wif + wash_wif. /* For most surveys: exceptions listed below. if (survey='1997 India') chorewif = cook_wif + wash_wif + clothwif. /* clothwif: India only. if (survey='2002 India') chorewif = cook_wif + wash_wif + clothwif. /* clothwif: India only. if (survey='2007 India') chorewif = cook_wif + wash_wif + clothwif. /* clothwif: India only. if (survey='2012 India') chorewif = cook_wif + wash_wif + clothwif. /* clothwif: India only. if (survey='2017 India') chorewif = cook_wif + wash_wif + clothwif. /* clothwif: India only. if (survey='1994 Brazil') chorewif = cook_wif + laundwif. /* BR 1994: included laundwif but not cleanwif. * Housework time in India 1992 is much lower than later India surveys, because fewer time-use questions asked in 1992; could make chorehus & chorewif missing in 1992. * Could exclude mind_hus & mind_wif from hwrk_hus & hwrk_wif, on the grounds that childminding can be done at the same time as other jobs. var label hwrk_HUS 'hours spent cooking/cleaning/laundry/shopping/childcare (per week): self/husband'. var label hwrk_WIF 'hours spent cooking/cleaning/laundry/shopping/childcare (per week): self/wife'. var label choreHUS 'hours spent cooking/cleaning/laundry (per week): self/husband'. var label choreWIF 'hours spent cooking/cleaning/laundry (per week): self/wife'. var label COOKpHUS 'hours preparing food before cooking (per week): self/husband'. var label COOKpWIF 'hours preparing food before cooking (per week): self/wife'. var label COOKcHUS 'hours cooking,excluding food preparation(per week): self/husb'. var label COOKcWIF 'hours cooking,excluding food preparation(per week): self/wife'. var label COOK_HUS 'hours spent cooking (per week): self/husband'. var label COOK_WIF 'hours spent cooking (per week): self/wife'. var label SHOP_HUS 'hours spent shopping (per week): self/husband'. var label SHOP_WIF 'hours spent shopping (per week): self/wife'. var label laundHUS 'hours spent cleaning clothes (per week): self/husband'. var label laundWIF 'hours spent cleaning clothes (per week): self/wife'. var label clothHUS 'hours making/mending clothes (per week): self/husband'. var label clothWIF 'hours making/mending clothes (per week): self/wife'. var label WASH_HUS 'hours spent on laundry/cleaning home(/day): self/husband'. var label WASH_WIF 'hours spent on laundry/cleaning home(/day): self/wife'. var label cleanHUS 'hours spent cleaning home (per week): self/husband'. var label cleanWIF 'hours spent cleaning home (per week): self/wife'. var label MIND_HUS 'hours spent minding children (per week): self/husband'. var label MIND_WIF 'hours spent minding children (per week): self/wife'. var label leisuHUS 'hours spent on leisure (per week): self/husband'. var label leisuWIF 'hours spent on leisure (per week): self/wife'. breakd hwrk_hus chorehus hwrk_wif chorewif by survey /cells mean. *@ ATTITUDES. if (sex=1) atWroles = AT_ROLES. if (sex=1) atMroles = ATpROLES. if (sex=0) atMroles = AT_ROLES. if (sex=0) atWroles = ATpROLES. var label AT_DIV_D 'Is heavy drinking sufficient for divorce?'. var label AT_DIV_V 'Is domestic violence sufficient for divorce?'. var label AT_DIV_U 'Is unfaithfulness sufficient for divorce?'. val label AT_DIV_D, AT_DIV_V, AT_DIV_U 1 'OK to divorce' 0 'not OK to divorce'. var label AT_jbSGL 'should an unmarried woman work outside home?'. var label AT_jbNoK 'should wife with no children work outside home?'. var label AT_jbInf 'should wife with preschool child work outside home?'. var label AT_jbKid 'should wife with schoolchild(ren) work outside home?'. recode AT_jbSGL AT_jbNOK AT_jbINF AT_jbKID (1=2)(2=1)(3=0)(4=sysmis). /* zzz does this apply to all surveys?. val label AT_jbSGL AT_jbNOK AT_jbINF AT_jbKID 2"full-time" 1"part-time" 0"not at all". var label at_h_out "Dis/agree: it's justified for a man to hit his wife, if she goes out without telling him". var label at_hNegl "Dis/agree: it's justified for a man to hit his wife, if she neglects the children". var label at_hArgu "Dis/agree: it's justified for a man to hit his wife, if she argues with him". var label at_hNSex "Dis/agree: it's justified for a man to hit his wife, if she refuses to have sex with him". var label at_hBurn "Dis/agree: it's justified for a man to hit his wife, if she burns the food". var label at_hCook "Dis/agree: it's justified for a man to hit his wife, if food is not cooked on time". var label at_hKeep "Dis/agree: it's justified for a man to hit his wife, if she doesn't give him her earnings". val lab at_h_out,at_hNegl,at_hArgu,at_hNSex,at_hBurn,at_hKeep at_hCook 1"yes" 0"no". * at_okHit is a Likert scale - so can't easily compare it with above zero/one variables. * 'Anti-feminist' statements: . var label at_ROLES "Dis/agree: Husb should earn/wife mind home". var label atpROLES "Spouse dis/agree: Husb should earn/wife mind home". var label atWroles "Woman/wife dis/agree: Husb should earn/wife mind home". var label atMroles "Man/husband dis/agree: Husb should earn/wife mind home". var label at_shame "Dis/agree: If a wife is unfaithful, it brings shame on her husband". var label at_okHit "Dis/agree: there are situations when it is justified for a man to hit his wife". var label at_pProb "Dis/agree: if a woman earns more than her husband, it will cause problems". var label at_eProb "Dis/agree: if a woman is more educated than her husband, it will cause problems". var label at_OBEY "Dis/agree: A wife should always obey her husband". var label at_CANok "Dis/agree: It's better if meals made from raw (not prepared) ingredients". var label at_boost "Dis/agree: If husband loses job, wife should try to boost his confidence". var label at_allow "Dis/agree: I choose to let my spouse make decisions for the family". var label at_wAlow "Dis/agree: women choose to let their spouse make decisions for the family". var label at_order "Dis/agree: I want to make household decisions myself, without having to discuss it with my spouse". var label at_fight "Dis/agree: it is shameful for a man to walk away from a fight". var label at_honor "Dis/agree: it is very important for a man to protect the honour of himself and his family". var label at_putUp "Dis/agree: A women should put up with a violent husband for her children". /* New in 2017. val lab at_roles,atpRoles,atMroles,atWroles,at_shame,at_okHIT,at_pProb,at_eProb,at_obey,at_CANok,at_boost,at_allow,at_wAlow,at_order,at_fight,at_honor, at_putUp 1'agree strongly' 2'agree' 3'neither agree/disagree' 4'disagree' 5'disagree strongly'. * Arguably, AT_allow & at_wAlow are vaguely feminist, if male respondent - depending on what decisions are meant (e.g. cooking?); * and at_letMe is antifeminist, if male respondent. * 'Pro-feminist' statements: . var label at_consi "Dis/agree: Husband should consider wife's opinion when making decision". var label at_wHAPY "Dis/agree: Woman happier if she has a job". var label at_oHAPY "Dis/agree: Family happier if woman has job". var label at_EqPay "Do you agree with law to give women equal pay to men?". var label at_QUICK "Dis/agree: Prefer meals which can be prepared quickly". var label at_USEFL "Dis/agree: Prepared food, e.g. cans, are very useful day-to-day". var label at_BUSY "Dis/agree: I am always short of time". var label at_PREPA "Dis/agree: Convenience foods can save time on cooking". var label at_decid "Dis/agree: Wife has the right to decide family expenditure". var label at_wJoin "Dis/agree: Wife may join social activity like religious gathering". var label at_w_say "Dis/agree: I wish women had more say in household decisions". var label at_letMe "Dis/agree: my spouse lets me make decisions for the family". var label at_noHit "Dis/agree: I am totally against violence". * New to Indonesia 2001-2 (related to feminism, but in complicated ways). var label at_nKids 'How many children do you wish to have?'. var label at_nSons 'How many sons do you wish to have?'. var label at_nDaug 'How many daughters do you wish to have?'. *var label wantKids 'If you could go back to the time when you had no children, how many children would you want?'. /* Use at_nKids instead. * New to Kenya 2004: . var label at_Hobey "Dis/agree: A husband should always obey his wife". var label at_chase "Dis/agree: During domestic violence, men often chase their wife away". * At this point, all Likert scale variables use 1'totally agree' to 5'totally disagree'. * Convert variables with "pro-feminist" statements, so they use {high-score=feminist}, for women respondents. recode at_consi,at_wHapy,at_oHapy,at_eqPay,at_quick,at_usefl,at_busy,at_prepa,at_decid,at_wJoin,at_w_say at_letMe at_noHit at_sEarn at_iEarn (1=5)(2=4)(3=3)(4=2)(5=1). value label at_consi,at_wHapy,at_oHapy,at_eqPay,at_quick,at_usefl,at_busy,at_prepa,at_decid,at_wJoin,at_w_say at_letMe at_noHit at_sEarn at_iEarn 5'agree strongly' 4'agree' 3'neither agree/disagree' 2'disagree' 1'disagree strongly'. * New to Nigeria 2005: . var label at_canDo "Is there anything a woman can do, to make husband spend less on alcohol etc?". var label at_CFhus "Do you [if male]/does your husband [if female] prefer car/drink/etc spending even if child hungry?". var label at_CFwif "Do you [if female]/does your wife [if male] prefer car/drink/etc spending even if child hungry?". *var label at_CFwif "Does your wife prefer car/drink/etc spending even if child hungry?". *var label at_CFhus "Does your husband prefer car/drink/etc spending even if child hungry?". var label at_CFmen "Do men prefer car/drink/etc spending even if child hungry?". var label at_cPray 'To cut his alcohol etc spending: she should pray/they should pray together'. var label at_cLove 'To cut his alcohol etc spending: show him respect/love/care, always be faithful'. var label at_cKids 'To cut his alcohol etc spending: talk to husband on importance of children'. var label at_cAdvi 'To cut his alcohol etc spending: advise; encourage him to save/be responsible'. var label at_cPlan 'To cut his alcohol etc spending: share financial decisions/buy essentials'. var label at_cEarn 'To cut his alcohol etc spending: she should support family financially'. val label at_canDo, at_CFwif,at_CFhus, at_CFmen at_cPray to at_cEarn 0'no' 1'yes'. * New to Chad 2008: . var label at_letMe "Dis/agree: 'My spouse lets me make decisions for the family' ". var label at_order "Dis/agree: 'I want to make decisions myself, without having to discuss it with my spouse' ". var label at_fight "Dis/agree: 'It is shameful for a man to walk away from a fight' ". var label at_noHit "Dis/agree: 'I am totally against violence' ". var label at_honor "Dis/agree: 'It is very important for a man to protect the honour of his family' ". * New to Cameroon 2009: . var label at_mKids "Dis/agree: 'A man should try to have as many children as he can' ". * New to Congo-Brazzaville in 2011: . var label at_sEarn "Dis/agree: 'A wife should earn money herself, and not totally depend on her husband’s income' ". var label at_iEarn "Dis/agree: 'I earn enough money to support everyone in this household' ". var label at_prote "Dis/agree: 'A wife usually obeys her husband in return for his protection' ". var label at_wantL "Dis/agree: 'I want to leave my husband due to GBV, but can't afford to' ". *@ RELIGION. recode religion (93,94=50). /*I used 93='Koli (India)' but www.encyclopedia.com/article-1G2-1839300328/kolis.html (2Jan2015) state "They are primarily Hindu". * I used 94='Nepali (India)'; 86% Nepalis in DHS NP4 Hindu; "Nepal was until 2006 the only kingdom in the world to be offically Hindu": https://books.google.co.uk/books?id=v2yiyLLOj88C. recode religion (15=13). /* I used 15 for 'Christa do reino de Deus', but http://en.wikipedia.org/wiki/Universal_Church_of_the_Kingdom_of_God claims they are Pentecostal. recode religion (90=99). /* This line is so I can use 95 for Mayan (in DHS surveys). val label religion 0 'No religion/atheist/agnostic' 10 'Christian (unspecified)' 11 'Protestant' 12 'Presbyterian' 13 'Evangelical/Wesleyan/Baptist/Pentecostal' 14 'Messianica' 16 '7th Day Adventist' 18 'Orthodox' 19 'Roman Catholic' 20 'Christian-related' 21 'Rosicrucian' 22 'Mormon' 23 "Jehovah's Witness" 24 "Rastafari" 30 'Muslim' 31 'Bahai' 40 'Jewish/Zionist' 50 'Hindu/Vaisnavist' 51 'Sikh' 52 'Kirat' 60 'Buddhist/neo-Buddhist' 61 'Jain' 70 'Parsi/Zoroastrian' 72 'Cao Dai' 75 'Confucian' 80 'Animist' 81 'Traditional African' 82 'Independent African' 83 'Candomble' 84 'Umbanda' 89 'Spiritualist/Voodoo' 91 'Soicho-Noye (Brazil)' 92 'Teluja (India)' 95 'Mayan' 99 'unspecified religion'. * Some values of religion (e.g. 31, 52, 72 and 75) aren't in WAS surveys; they are included for compatibility with DHS surveys. I can't tell what "Teluja" means. * 'Donyi polo' is a type of animism in Andhra Pradesh, and 'Sanamahi' a kind of Hinduism, according to Wikipedia (16th May 2010). *crosstab religion by live_cit. *@ FAMILY STRUCTURE. if (fam_NGEN>2) fam_TYPE=5. if (famRELAT=0 and famINLAW=0) fam_TYPE=1. if (sex=0 and famRELAT=1 & famINLAW=0) fam_TYPE=2. if (sex=1 and famRELAT=1 & famINLAW=0) fam_TYPE=3. if (sex=0 and famRELAT=0 & famINLAW=1) fam_TYPE=3. if (sex=0 and famRELAT=0 & famINLAW=1) fam_TYPE=2. if (famRELAT=1 and famINLAW=1) fam_TYPE=4. recode fam_TYPE (5=sysmis). /*One HH in BM92: not clear. var label fam_TYPE 'is the family nuclear, or extended?'. val label fam_TYPE 1"nuclear/single person" 2"relatives of husband" 3"relatives of wife" 4"both husb & wife relatives". var label fam_NGEN 'number of generations living in the HH'. val label fam_NGEN 1"one generation in HH" 2"2 generations in HH" 3"3 generations in HH" 4"4 generations in HH" 5"5 generations in HH". var label famINlaw 'Does the resp have any in-laws in the HH?'. val label famINlaw 0"no in-laws in hhold" 1"in-law(s) in hhold". var label famRELAT 'Does the resp have blood relatives in the HH?'. val label famRELAT 0"no relative in hhold" 1"relative(s) in hhold". * For most WAS surveys, I know the age of each hhold member, but not WAS ZA. if (INT_year=2000) n_age_0 = mAGE_0 + fAGE_0. if (INT_year=2000) n_age_1 = mAGE_1 + fAGE_1. if (INT_year=2000) n_age_2 = mAGE_2 + fAGE_2. if (INT_year=2000) n_age_3 = mAGE_3 + fAGE_3. *if(INT_year=2000) n_age_4 = ((mAGE_4_6 + fAGE_4_6) / 3). *if(INT_year=2000) n_age_5 = ((mAGE_4_6 + fAGE_4_6) / 3). *if(INT_year=2000) n_age_6 = ((mAGE_4_6 + fAGE_4_6) / 3). *if(INT_year=2000) n_age_7 = ((mAGE_7_9 + fAGE_7_9) / 3). *if(INT_year=2000) n_age_8 = ((mAGE_7_9 + fAGE_7_9) / 3). *if(INT_year=2000) n_age_9 = ((mAGE_7_9 + fAGE_7_9) / 3). *if(INT_year=2000) n_age_10 = ((mAGE1012 + fAGE1012) / 3). *if(INT_year=2000) n_age_11 = ((mAGE1012 + fAGE1012) / 3). *if(INT_year=2000) n_age_12 = ((mAGE1012 + fAGE1012) / 3). *if(INT_year=2000) n_age_13 = ((mAGE1314 + fAGE1314) / 2). *if(INT_year=2000) n_age_14 = ((mAGE1314 + fAGE1314) / 2). *if(INT_year=2000) n_age_15 = mAGE15 + fAGE15. *if(INT_year=2000) n_age_16 = ((mAGE1618 + fAGE1618) / 3). *if(INT_year=2000) n_age_17 = ((mAGE1618 + fAGE1618) / 3). *if(INT_year=2000) n_age_18 = ((mAGE1618 + fAGE1618) / 3). * Ensure the number in each age-group is a whole number. if (INT_year=2000) n_age_4 = trunc((mAGE_4_6 + fAGE_4_6) / 3). if (INT_year=2000) n_age_6 = trunc((mAGE_4_6 + fAGE_4_6) / 3). if (INT_year=2000) n_age_5 = (mAGE_4_6 + fAGE_4_6) - n_age_4 - n_age_6. if (INT_year=2000) n_age_7 = trunc((mAGE_7_9 + fAGE_7_9) / 3). if (INT_year=2000) n_age_9 = trunc((mAGE_7_9 + fAGE_7_9) / 3). if (INT_year=2000) n_age_8 = (mAGE_7_9 + fAGE_7_9) - n_age_7 - n_age_9. if (INT_year=2000) n_age_10 = trunc((mAGE1012 + fAGE1012) / 3). if (INT_year=2000) n_age_12 = trunc((mAGE1012 + fAGE1012) / 3). if (INT_year=2000) n_age_11 = (mAGE1012 + fAGE1012) - n_age_10 - n_age_12. if (INT_year=2000) n_age_13 = trunc((mAGE1314 + fAGE1314) / 2). if (INT_year=2000) n_age_14 = (mAGE1314 + fAGE1314) - n_age_13. if (INT_year=2000) n_age_15 = mAGE15 + fAGE15. if (INT_year=2000) n_age_16 = trunc((mAGE1618 + fAGE1618) / 3). if (INT_year=2000) n_age_18 = trunc((mAGE1618 + fAGE1618) / 3). if (INT_year=2000) n_age_17 = (mAGE1618 + fAGE1618) - n_age_16 - n_age_18. * The above syntax often puts two of the 4-to-6 agegroup as age 5, but none age 4 or 6; correct this. if (INT_year=2000 & n_age_5>1 & n_age_4=0) n_age_45 = 2. if (INT_year=2000 & n_age_45 = 2) n_age_5 = n_age_5 - 1. if (INT_year=2000 & n_age_45 = 2) n_age_4 = n_age_4 + 1. * Similarly for ages 7 to 9. if (INT_year=2000 & n_age_8>1 & n_age_7=0) n_age_78 = 2. if (INT_year=2000 & n_age_78 = 2) n_age_8 = n_age_8 - 1. if (INT_year=2000 & n_age_78 = 2) n_age_7 = n_age_7 + 1. * Similarly for ages 10 to 12. if (INT_year=2000 & n_age_11>1 & n_age_10=0) n_age_1011=2. if (INT_year=2000 & n_age_1011=2) n_age_11 = n_age_11 - 1. if (INT_year=2000 & n_age_1011=2) n_age_10 = n_age_10 + 1. *if (INT_year=2000 & n_age_14>1 & n_age_13=0) n_age_1314=2. /* No cases of n_age_1314=2. * Similarly for ages 16 to 18. if (INT_year=2000 & n_age_17>1 & n_age_16=0) n_age_1617=2. if (INT_year=2000 & n_age_1617=2) n_age_17 = n_age_17 - 1. if (INT_year=2000 & n_age_1617=2) n_age_16 = n_age_16 + 1. *. if (INT_year=2000) n_age_dk = 0. *cros survey by n_age_dk by sourcefile. * Could define 'adult' as over 18 (arbitrary). *compute N_CHILDR = mAGE_0 + mAGE_1 + mAGE_2 + mAGE_3 + mAGE_4_6+mAGE_7_9+ mAGE1012+ mAGE1314+ mAGE15+ mAGE1618 + fAGE_0 + fAGE_1 + fAGE_2 + fAGE_3 + fAGE_4_6+ fAGE_7_9+ fAGE1012+ fAGE1314+ fAGE15+ fAGE1618. *compute N_ADULTS = mAGE1924+ mAGE2534 + mAGE35__ + fAGE1924+ fAGE2534+ fAGE35__ . * Could define 'adult' as age 16+ (arbitrary): I changed to this definition on 26th September 2009 (from 19+). *compute N_CHILDR = mAGE_0 + mAGE_1 + mAGE_2 + mAGE_3 + mAGE_4_6+mAGE_7_9 + mAGE1012+ mAGE1314+ mAGE15 + fAGE_0 + fAGE_1 + fAGE_2 + fAGE_3 + fAGE_4_6+ fAGE_7_9+ fAGE1012+ fAGE1314+ fAGE15. *compute N_ADULTS = mAGE1618 + fAGE1618 + mAGE1924 + mAGE2534 + mAGE35__ + fAGE1924+ fAGE2534+ fAGE35__ . * From 27th September 2009, define adult as age 18 or older. compute n_childr =n_age_0+n_age_1+n_age_2+n_age_3+n_age_4+n_age_5+n_age_6+n_age_7+n_age_8+n_age_9+n_age_10+n_age_11+n_age_12+n_age_13 +n_age_14+n_age_15+n_age_16+n_age_17. compute n_adults=n_age_18 + mAGE1924+ mAGE2534 + mAGE35__ + fAGE1924 + fAGE2534 + fAGE35__ + n_age_dk. /* Assume if unknown age, is adult. compute n_male=mAGE_0+mAGE_1+mAGE_2+mAGE_3+mAGE_4_6+mAGE_7_9+mAGE1012+mAGE1314+mAGE15+mAGE1618+mAGE1924+mAGE2534 +mAGE35__ . compute n_female=fAGE_0+ fAGE_1+ fAGE_2+ fAGE_3+ fAGE_4_6+ fAGE_7_9+ fAGE1012+fAGE1314+ fAGE15 +fAGE1618 +fAGE1924+ fAGE2534+fAGE35__ . compute n_PEOPLE = N_ADULTS + N_CHILDR. compute n_BABIES = mAGE_0 +fAGE_0 +mAGE_1+fAGE_1 +mAGE_2 +fAGE_2 +mAGE_3 +fAGE_3. var label N_CHILDR 'Number of children (age under 19) in the household'. var label N_ADULTS 'Number of adults (age 19 or more) in the household'. var label N_PEOPLE 'Number of persons in the household'. var label N_MALE 'Number of males (boys & men) in the household'. var label N_FEMALE 'Number of females (girls & women) in the household'. var label N_EARNER 'Number of people employed in the household'. var label mAge_0 'Number of males in household, age 0'. var label mAge_1 'Number of males in household, age 1'. var label mAge_2 'Number of males in household, age 2'. var label mAge_3 'Number of males in household, age 3'. var label mAge_4_6 'Number of males in household, age 4 to 6'. var label mAge_7_9 'Number of males in household, age 7 to 9'. var label mAge1012 'Number of males in household, age 10 to 12'. var label mAge1314 'Number of males in household, age 13 to 14'. var label mAge15 'Number of males in household, age 15'. var label mAge1618 'Number of males in household, age 16 to 18'. var label mAge1924 'Number of males in household, age 19 to 24'. var label mAge2534 'Number of males in household, age 25 to 34'. var label mAge35__ 'Number of males in household, age 35 or more'. var label fAge_0 'Number of females in household, age 0'. var label fAge_1 'Number of females in household, age 1'. var label fAge_2 'Number of females in household, age 2'. var label fAge_3 'Number of females in household, age 3'. var label fAge_4_6 'Number of females in household, age 4 to 6'. var label fAge_7_9 'Number of females in household, age 7 to 9'. var label fAge1012 'Number of females in household, age 10 to 12'. var label fAge1314 'Number of females in household, age 13 to 14'. var label fAge15 'Number of females in household, age 15'. var label fAge1618 'Number of females in household, age 16 to 18'. var label fAge1924 'Number of females in household, age 19 to 24'. var label fAge2534 'Number of females in household, age 25 to 34'. var label fAge35__ 'Number of females in household, age 35 or more'. var label n_BABIES 'Number of children aged 0 to 3 in the HH'. val label n_BABIES 0"no child age under 4 in HHold". * For BM92, I created n_INFANTS (age 4 to 5); n_JUNIOR (age 6 to 11); and n_SCHOOL (12 to 17). * But don't, now (they can't be derived from mAGE_4_6+AGE_4_6 etc; & my agebands are arbitrary). * HH_equiv: my adaptation of methodology in Department of Social Security (1993), 'Households below average income: a statistical analysis 1979-1990/91'; * I use their "after housing costs" version, because housing costs are low in most "Third World" households. * I use different agebands (e.g. age 4-6 rather 5-7) to DSS, because South African WAS data only has these agebands (age 4-6 etc). if (N_ADULTS = 0) HH_EQUIV = 0 . /* This line added 31st Oct 2008. if (N_ADULTS = 1) HH_EQUIV = 0.55 . if (N_ADULTS = 2) HH_EQUIV = 1.00 . if (N_ADULTS = 3) HH_EQUIV = 1.45 . if (N_ADULTS > 3) HH_EQUIV = 1.45 + ( 0.40 *( N_ADULTS - 3 ) ). *compute HH_EQUIV = HH_EQUIV + ( 0.07 *( mAge_0 + mAge_1 + fAge_0 + fAge_1 ) ). *compute HH_EQUIV = HH_EQUIV + ( 0.18 *( mAge_2 + mAge_3 + fAge_2 + fAge_3 ) ). *compute HH_EQUIV = HH_EQUIV + ( 0.21 *( mAge_4_6 + fAge_4_6 ) ). *compute HH_EQUIV = HH_EQUIV + ( 0.23 *( mAge_7_9 + fAge_7_9 ) ). *compute HH_EQUIV = HH_EQUIV + ( 0.26 *( mAge1012 + fAge1012 ) ). *compute HH_EQUIV = HH_EQUIV + ( 0.28 *( mAge1314 +mAge15 + fAge1314 +fAge15 ) ). *compute HH_EQUIV = HH_EQUIV + ( 0.38 *( mAge1618 + fAge1618 ) ). * The above 7 lines were used until 27th September 2009; version below is slightly closer to the DSS definition. compute HH_EQUIV = HH_EQUIV + ( 0.07 *( n_age_0 + n_age_1 ) ). compute HH_EQUIV = HH_EQUIV + ( 0.18 *( n_age_2 + n_age_3 + n_age_4 ) ). compute HH_EQUIV = HH_EQUIV + ( 0.21 *( n_age_5 + n_age_6 + n_age_7 ) ). compute HH_EQUIV = HH_EQUIV + ( 0.23 *( n_age_8 + n_age_9 + n_age_10 ) ). compute HH_EQUIV = HH_EQUIV + ( 0.26 *( n_age_11 + n_age_12 ) ). compute HH_EQUIV = HH_EQUIV + ( 0.28 *( n_age_13 + n_age_14 + n_age_15 ) ). compute HH_EQUIV = HH_EQUIV + ( 0.38 *( n_age_16 + n_age_17 ) ). var label HH_EQUIV 'Household equivalence scale (calculated to create HHincEQI)'. *freq HH_EQUIV. /* There shouldn't be any zero values; but 50 cases are. recode HH_EQUIV (0 = sysmis). if (HHincome>0) HHincEQI = HHincome / HH_EQUIV. *breakd n_age_0 to n_age_18, n_adults by survey /cells mean. * Work out calorie requirements, using info from WorldBank: www.worldbank.org/html/prdph/lsms/country/pan97/docs/poveng.pdf, downloaded 15/11/2003, p.3. * Figures below assume "moderate activity"; employed people use more calories, so could use job. Male children, then female children, then adults. *compute ca_requi = (mage_0 * 738 ) + (mage_1 * 1100 ) + (mage_2 * 1300 ) + (mage_3 * 1500 ) + (mage_4_6 * (1500+1750+1750)/3 ) + (mage_7_9 * 2000 ) + (mage1012 * (2200+2200+2350)/3 ) + (mage1314 * (2350+2650)/2 ) + (mage15 * 2650 ) + (fage_0 * 738 ) + (fage_1 * 1100 ) + (fage_2 * 1300 ) + (fage_3 * 1500 ) + (fage_4_6 * (1500+1600+1600)/3 ) + (fage_7_9 * 1700 ) + (fage1012 * (1900+1900+2000)/3 ) + (fage1314 * (2000+2100)/2 ) + (fage15 * 2100 ) + (n_adults * (3100+2100)/2 ). *Alternative version: www.nutrition.org.uk/information/energyandnutrients/energy.html downloaded 21 Nov 2003. *compute ca_requi = (age_0 ) *(545+515+690+645+825+765+920+865)/8 + (age_1 + age_2 + age_3 ) *(1230+1165)/2 + (age_4 + age_5 + age_6 ) *(1715+1545)/2 + (age_7 + age_8 + age_9 + age10) *(1970+1740)/2 + (age11 + age12 + age13 + age14) *(2220+1845)/2 + (age15 ) *(2755+2110)/2 + (n_adults ) *(2550+1940)/2. * Now more detailed versions, to assess if a wife can afford to leave her husband - assume she'd take young children with her; how old?. *compute pNrelat = p1_relat. /* Could only include (female) respondent's children, in list of people to take. *. recode age_hus, age_wif (sysmis=999). /* Temporary: to reduce missing data, assume husb & wife age at least 24 (for ca_requi_u24). compute ca_requi_u24 = 0. compute ca_requi_u21 = 0. compute ca_requi_u18 = 0. compute ca_requi_u16 = 0. compute ca_requi = 0. compute proRequi = 0. *. compute pNsex = 1. compute pNage = age_wif. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. if (age_wif> 0) ca_requi_u24 = cal_need. /* Add calories for children below. if (age_wif> 0) ca_requi_u21 = cal_need. /* Add calories for children below. if (age_wif> 0) ca_requi_u18 = cal_need. /* Add calories for children below. if (age_wif> 0) ca_requi_u16 = cal_need. /* Add calories for children below. if (age_wif<24) ca_requi_u24 = 0. if (age_wif<21) ca_requi_u21 = 0. if (age_wif<18) ca_requi_u18 = 0. if (age_wif<16) ca_requi_u16 = 0. /* If wife is young, don't count her calorie needs twice. compute ca_requi = 0. *. compute pNsex = p1_sex. compute pNage = p1_age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p2_sex. compute pNage = p2_age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p3_sex. compute pNage = p3_age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p4_sex. compute pNage = p4_age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p5_sex. compute pNage = p5_age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p6_sex. compute pNage = p6_age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p7_sex. compute pNage = p7_age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p8_sex. compute pNage = p8_age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p9_sex. compute pNage = p9_age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p10sex. compute pNage = p10age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p11sex. compute pNage = p11age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p12sex. compute pNage = p12age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p13sex. compute pNage = p13age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p14sex. compute pNage = p14age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p15sex. compute pNage = p15age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p16sex. compute pNage = p16age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p17sex. compute pNage = p17age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p18sex. compute pNage = p18age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p19sex. compute pNage = p19age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p20sex. compute pNage = p20age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p21sex. compute pNage = p21age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p22sex. compute pNage = p22age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. compute pNsex = p23sex. compute pNage = p23age. include 'D:\_WAS\SPSS syntax for all years\calculate calorie need of 1 person.sps'. *. *if (survey = '2000 S. Africa') ca_requi = (mage_0 * 645.93 ) + (mage_1 * 956.93 ) + (mage_2 * 1148.32 ) + (mage_3 * 1339.71 ) + (mage_4_6 * (1411.48+1483.25+1578.94)/3 ) + (mage_7_9 * (1674.64+1746.41+1866.02)/3 ) + (mage1012 * (1985.64+2105.26+2224.88)/3 ) + (mage1314 * (2392.34+2535.88)/2 ) + (mage15 * 2679.42 ) + (fage_0 * 598.08 ) + (fage_1 * 909.09 ) + (fage_2 * 1100.47 ) + (fage_3 * 1267.94 ) + (fage_4_6 * (1315.78+1363.63+1459.33)/3 ) + (fage_7_9 * (1555.02+1650.71+1746.41)/3 ) + (fage1012 * (1818.18+1913.87+2033.49)/3 ) + (fage1314 * (2129.18+2200.95)/2 ) + (fage15 * 2248.8 ) + (n_adults * (2822.96+2703.34+2488.03+2272.72 + 2296.65+2177.03+2081.33+1985.64)/8). * Number aged 16-18: cal_need=2822.96, 2918.66, 2990.43; 2272.72, 2296.65, 2320.57 . *if (survey = '2000 S. Africa') ca_requi_u16= (mage_0 * 645.93 ) + (mage_1 * 956.93 ) + (mage_2 * 1148.32 ) + (mage_3 * 1339.71 ) + (mage_4_6 * (1411.48+1483.25+1578.94)/3 ) + (mage_7_9 * (1674.64+1746.41+1866.02)/3 ) + (mage1012 * (1985.64+2105.26+2224.88)/3 ) + (mage1314 * (2392.34+2535.88)/2 ) + (mage15 * 2679.42 ) + (fage_0 * 598.08 ) + (fage_1 * 909.09 ) + (fage_2 * 1100.47 ) + (fage_3 * 1267.94 ) + (fage_4_6 * (1315.78+1363.63+1459.33)/3 ) + (fage_7_9 * (1555.02+1650.71+1746.41)/3 ) + (fage1012 * (1818.18+1913.87+2033.49)/3 ) + (fage1314 * (2129.18+2200.95)/2 ) + (fage15 * 2248.8 ) + ( (2296.65+2177.03+2081.33+1985.64)/4 ). /* Assume mum is an adult, but of unknown age (could I use age_wif?). *if (survey = '2000 S. Africa') ca_requi_u18 = ca_requi_u16. /* ca_requi_u16 is correct; ca_requi_u18 is underestimated. *if (survey = '2000 S. Africa') ca_requi_u21 = ca_requi_u16. /* ca_requi_u16 is correct; ca_requi_u21 is underestimated. *if (survey = '2000 S. Africa') ca_requi_u24 = ca_requi_u16. /* ca_requi_u16 is correct; ca_requi_u24 is underestimated. * Protein requirements, from http://whqlibdoc.who.int/trs/WHO_TRS_935_eng.pdf table 47 (=p.244), downloaded 26 Sept 2009: . *if (survey = '2000 S. Africa') proRequi = (mage_0 * 10.2 ) + (mage_1 * 11.6 ) + (mage_2 * 11.9 ) + (mage_3 * 13.1 ) + (mage_4_6 * 17.1 ) + (mage_7_9 * 25.9 ) + (mage1012 * 40.5 ) + (mage1314 * 40.5 ) + (mage15 * 57.9 ) + (fage_0 * 9.4 ) + (fage_1 * 10.8 ) + (fage_2 * 11.4 ) + (fage_3 * 12.7 ) + (fage_4_6 * 16.2 ) + (fage_7_9 * 26.2 ) + (fage1012 * 41.0 ) + (fage1314 * 41.0 ) + (fage15 * 47.4 ) + (n_adults * (57.9 + 47.4)/2). * Above block is an approximation to proRequi (in ZA). * Alternative approach: Assume boys & girls have similar protein requirements. * Protein requirements, from http://whqlibdoc.who.int/trs/WHO_TRS_935_eng.pdf table 47 (=p.244), downloaded 26 Sept 2009: . *compute proRequi = (n_age_0 * (10.2 + 9.4 )/2 ) + (n_age_1 * (11.6 + 10.8 )/2 ) + (n_age_2 * (11.9 + 11.4 )/2 ) + (n_age_3 * (13.1 + 12.7 )/2 ) + (n_age_4 * (17.1 + 16.2 )/2 ) + (n_age_5 * (17.1 + 16.2 )/2 ) + (n_age_6 * (17.1 + 16.2 )/2 ) + (n_age_7 * (25.9 + 26.2 )/2 ) + (n_age_8 * (25.9 + 26.2 )/2 ) + (n_age_9 * (25.9 + 26.2 )/2 ) + (n_age_10* (25.9 + 26.2 )/2 ) + (n_age_11* (40.5 + 41.0 )/2 ) + (n_age_12* (40.5 + 41.0 )/2 ) + (n_age_13* (40.5 + 41.0 )/2 ) + (n_age_14* (40.5 + 41.0 )/2 ) + (n_age_15* (57.9 + 47.1 )/2 ) + (n_adults * (57.9 + 47.4 )/2 ). if (pNsex=0 & pNage= 0 ) pro_need= 10.2 . if (pNsex=0 & pNage= 1 ) pro_need= 11.6 . if (pNsex=0 & pNage= 2 ) pro_need= 11.9 . if (pNsex=0 & pNage= 3 ) pro_need= 13.1 . if (pNsex=0 & (pNage> 3 & pNage< 7) ) pro_need= 17.1 . if (pNsex=0 & (pNage> 6 & pNage<11) ) pro_need= 25.9 . if (pNsex=0 & (pNage>10 & pNage<15) ) pro_need= 40.5 . if (pNsex=0 & (pNage>14 & pNage<19) ) pro_need= 57.9 . if (pNsex=0 & pNage>18 ) pro_need= 57.9 . /* I can't see a figure for adults, so I assume same as 15-18. if (pNsex=1 & pNage= 0 ) pro_need= 9.4 . if (pNsex=1 & pNage= 1 ) pro_need= 10.8 . if (pNsex=1 & pNage= 2 ) pro_need= 11.4 . if (pNsex=1 & pNage= 3 ) pro_need= 12.7 . if (pNsex=1 & (pNage> 3 & pNage< 7) ) pro_need= 16.2 . if (pNsex=1 & (pNage> 6 & pNage<11) ) pro_need= 26.2 . if (pNsex=1 & (pNage>10 & pNage<15) ) pro_need= 41.0 . if (pNsex=1 & (pNage>14 & pNage<19) ) pro_need= 47.4 . if (pNsex=1 & pNage>18 ) pro_need= 47.4 . /* I can't see a figure for adults, so I assume same as 15-18. *. *freq ca_requi. /* 45 cases of (ca_requi=0) - remove them. *desc ca_requi ca_requi_u24 ca_requi_u21 ca_requi_u18 ca_requi_u16. recode ca_requi ca_requi_u24 ca_requi_u21 ca_requi_u18 ca_requi_u16 (0=sysmis). var lab ca_requi 'Calories required: estimated from number & ages of household members'. var lab ca_requi_u24 'Calories required: mum, and children in household age under 24'. var lab ca_requi_u21 'Calories required: mum, and children in household age under 21'. var lab ca_requi_u18 'Calories required: mum, and children in household age under 18'. var lab ca_requi_u16 'Calories required: mum, and children in household age under 16'. var lab proRequi 'Protein required: estimated from number & ages of household members'. *breakd ca_requi_u16 by survey /cells mean count. /* This line useful to compare ZA with other countries. *breakd ca_requi_u18 by survey /cells mean count. /* This line useful to compare ZA with other countries. *breakd ca_requi_u21 by survey /cells mean count. /* This line useful to compare ZA with other countries. *breakd ca_requi_u24 by survey /cells mean count. /* This line useful to compare ZA with other countries. recode age_hus, age_wif (999 = sysmis). *@ HOUSEHOLD FINANCIAL MANAGEMENT. * SYSTEM TO IDENTIFY RELATION TO RESPONDENT: uses 3 dimensions. * First digit is generation (-100=child; +100=parent; etc); * second digit is sibling relationship ( 010=sibling=brother/sister, etc); * the third digit is spouse ( 001=respondent's spouse, etc). *Codes 801-803 are for mindKids; 2,3,4,etc for managMon,finalSay,at_final; other codes for p1_relat...p15relat. value label p1_relat,p2_relat,p3_relat,p4_relat,p5_relat,p6_relat,p7_relat,p8_relat,p9_relat,p10relat, p11relat,p12relat,p13relat,p14relat,p15relat,p16relat,p17relat,p18relat,p19relat,p20relat, p21relat,p22relat,p23relat MANAGmon,FINALsay,AT_final,MINDkids HEALTsay,VISITsay,FOOD_say,CHILDsay finalH_m,finalW_m 200"Grandparent" 210"Great-aunt/uncle" 201"Grandparent of spouse" 100"Parent" 110"Aunt/uncle" 101"Parent of spouse" 111"Aunt/uncle of spouse" 102"Parent's spouse (step-parent)" 000"Respondent" 010"Sibling" 030"Cousin" 001"Respondent's spouse" 011"Sibling's spouse" 021"Spouse's sibling" 012"Brother/sister-in-law" -100"Child" -103"adopted child" -110"Nephew/neice" -101"Child of spouse" -121"Nephew/neice of spouse" -102"Child's spouse" -200"Grandchild" -210"Great-nephew/neice" -201"Grandchild of spouse" -221"Spouse's sib's grandchild" 991 "head-of-household" 992 "main earner" -300"Great-grandchild" 993 "all in household" 994 "self & other(s), e.g. self & child" 995 "non-relative, e.g. friend or lodger" 996 "other(s): unspecified" 997 "other in-law; or spouse & in-law(s)" 998 "other relative(s)" 999 "elders, e.g. parents" 1.1 "husband", 1.2 "wife" 2 "husband & wife together" 3 "husband & wife separately" 4 "husband's 2nd wife" 5 "other, not just husband & wife" 801 "child's parents" 802 "no children in hh" 803 "maid/baby-sitter" -4 "Not in ID 2002 (data entry error)" -3 "no earnings". * If respondent is male, 'self' means man/husband; if female, 'self' is woman/wife. if (MANAGmon =0 & sex=0) MANAGmon = 1.1 . if (MANAGmon =0 & sex=1) MANAGmon = 1.2 . if (MANAGmon =1 & sex=0) MANAGmon = 1.2 . if (MANAGmon =1 & sex=1) MANAGmon = 1.1 . if (FINALsay =0 & sex=0) FINALsay = 1.1 . if (FINALsay =0 & sex=1) FINALsay = 1.2 . if (FINALsay =1 & sex=0) FINALsay = 1.2 . if (FINALsay =1 & sex=1) FINALsay = 1.1 . if (AT_FINAL =0 & sex=0) AT_FINAL = 1.1 . if (AT_FINAL =0 & sex=1) AT_FINAL = 1.2 . if (AT_FINAL =1 & sex=0) AT_FINAL = 1.2 . if (AT_FINAL =1 & sex=1) AT_FINAL = 1.1 . * If managMon is 'husb & wife separately' or 'husb & wife jointly' [or 'other'?] but respondent is single/widowed, treat as 'single husb' or 'single wife'. * However, only change this if there is just 1 adult in the household (the respondent may not be the head-of-household). *crosstabs managMon by married. if (managMon=2 & sex=0 & (married=3 or married=4) & N_adults=1) managMon = 1.1. if (managMon=2 & sex=1 & (married=3 or married=4) & N_adults=1) managMon = 1.2. if (managMon=3 & sex=0 & (married=3 or married=4) & N_adults=1) managMon = 1.1. if (managMon=3 & sex=1 & (married=3 or married=4) & N_adults=1) managMon = 1.2. if (INT_year<>1994) MANAGm_s = MANAGmon. if (INT_year<>1994) FINALs_s = FINALsay. if (INT_year<>1994) AT_FIN_s = AT_final. if (INT_year<>1994) MINDki_s = MINDkids. recode MANAGm_s, FINALs_s, AT_FIN_s, MINDki_s (1.1=0)(1.2, 4=1)(2, 3, 91=2)(else= -9). val label MANAGm_s, FINALs_s, AT_FIN_s, MINDki_s 0"male" 1"female" 2"both males & females". var label MANAGmon 'Person who organises household money'. var label FINALsay 'Person with final say over household money'. var label AT_FINAL 'Who should have final say over HH money?'. var label MANAGm_s 'Sex of person organising household money'. var label FINALs_s 'Sex of person with final say over HH money'. var label AT_FIN_s 'Sex of who should have final say: HH money'. var label MINDki_s 'Sex of who minds children in the household'. var label bankResp "Does the respondent have a bank account (own, or joint account)?". var label bankSpou "Does the respondent's spouse have a bank account (own, or joint)?". var label bank_hus "Does the husband have a bank account (own, or joint account)?". var label bank_wif "Does the wife have a bank account (own, or joint account)?". var label bank_jnt "Do respondent & spouse have a joint bank account?". value label bankResp bankSpou bank_hus bank_wif bank_jnt 1'has account(s)' 0'does not have an account'. * CONTROL asked in 2000 for the first time; the precision (number of values used) varies between countries. var label CONTROL "How much control do you have in household decisions?". value label CONTROL 0 'no control' 15 'little control' 25 'some control' 35 'some control' 50 'equal control' 65 'a lot of control' 75 'most of the control' 85 'most of the control' 100 'complete control'. *values 25 & 75 are for IN & NG & EG only; 15, 35, 65 & 85 only for other WAS surveys. * New to Egypt 2006: . value labels TALKwork TALKplan TALKkids TALKmony TALKcomu 100"regularly" 50"sometimes" 0"never" -1"don't know" -2"no answer" -9"missing". var label TALKwork "Does your spouse discuss work with you?". var label TALKplan "Does your spouse discuss future plans with you?". var label TALKkids "Does your spouse discuss your children's activities with you?". var label TALKmony "Does your spouse discuss money with you?". var label TALKcomu "Does your spouse discuss things that happen in the community with you?". * Avoid leaving violence as missing data, if there were no arguments. recode violence argument (sysmis= -33). if (argument= 0 & violence < 0) violence = 0. recode violence argument (-33=sysmis). *cros violence by argument. if (viol_hit =0) violence = 0. if (violHitB =0) violence = 0. if (viol_hit >0) violence = 1. if (violHitB >0) violence = 1. if (sex=1) viol_hus = viol_hit. if (sex=0) viol_hus = violHitB. if (sex=0) viol_wif = viol_hit. if (sex=1) viol_wif = violHitB. recode viol_wif viol_hus violence (-9 = sysmis). /* This line is for India 2017. if (sex=1 & survey<>'2005-6 Egypt') vioR_hus = vioM_hit. if (sex=0 & survey<>'2005-6 Egypt') vioR_hus = vioMHitB. if (sex=0 & survey<>'2005-6 Egypt') vioR_wif = vioM_hit. if (sex=1 & survey<>'2005-6 Egypt') vioR_wif = vioMHitB. * vioR (EG only) is similar to vioM (Chad and Cameroon); but they're not the same - EG is 6 months. *. if (sex=0) violBhus = violBeat. if (sex=1) violBwif = violBeat. *. var label argument 'Have you ever had an argument with your spouse?'. var label violence 'Have these arguments led to violence between you?'. var label viol_hit 'Have you hit your partner?'. var label violHitB 'Have you been hit by your partner?'. var label vioMhitB 'Have you been hit by your partner, in the previous month?'. var label viol_hus 'Male respondent (or partner of female resp) experienced violence'. var label viol_wif 'Female respondent (or partner of male resp) experienced violence'. var label vioR_hus 'Male respondent (or partner of female resp) experienced violence in last month/6 months'. var label vioR_wif 'Female respondent (or partner of male resp) experienced violence in last month/6 months'. var label violBhus 'Male respondent (or partner of female resp) has been beaten'. var label violBwif 'Female respondent (or partner of male res) has been beaten'. val label argument,violence viol_hit,violHitB viol_hus,viol_wif vioR_wif violBhus,violBwif violBeat violHumi violThre violPush violSlap violPunc violKick violBurn violTkni violKnif violRape 0'no' 0.5'yes: sometimes' 1'yes' 7.5'yes: often'. /* Some of these value labels aren't used in Kenya. var label BusinWho 'Who is responsible for the family business?'. *@ SPENDING. *I add a few variables together below (e.g. sp_oilBr = sp_oil + sp_bread) to allow comparisons between WAS surveys. if (INT_year = 1992 or INT_year=1997 or INT_year >2002) sp_oilBr = sp_oil + sp_bread. if (INT_year<>1994) sp_br_mi = sp_bread + sp_milk. /*Only works on 2004. if (INT_year = 1994) sp_bDair = sp_br_mi + sp_chees. if (INT_year<>1994 & INT_year< 2001) sp_bDair = sp_bread + sp_dairy. if (survey<>'2005-6 Egypt') sp_cereB = sp_cerea + sp_bread. if (INT_year < 2005 or INT_year = 2007) sp_mSnak = sp_fMeal + sp_snack. if (INT_year<>1992 & INT_year<>1997 & sp_drink>-1 & sp_tobac>-1) sp_CigDr = sp_drink + sp_tobac. if (survey<>'2005-6 Egypt') sp_leisT = sp_leisu + sp_cigDr. * sp_leisu includes cigarettes in EG, but not in BR and ID WAS surveys; add sp_leisu+sp_tobac (in WAS IN surveys), for comparability. if (sp_laund> -1) & (sp_maid>-1) sp_hHelp = sp_laund + sp_maid. * WAS Egypt combines 'utility' bills together: it mentions electricity water & phone on the questionnaire, and I assume it also includes gas bills. *breakd sp_leisu sp_leisT sp_cigdr sp_tobac sp_drink by survey/cells mean. *breakd sp_el_wa sp_elect sp_gas sp_phone sp_energ sp_water sp_utili by survey/cells mean. if (INT_year = 2004) sp_el_wa = sp_elect + sp_water. if (INT_year = 2004 or INT_year = 2007 or survey='2002 India' or survey='2012 India' or survey='2017 India') sp_utili = sp_el_wa + sp_gas + sp_phone. * It would be possible to combine water+electricity+gas (without phone) as 'utilities': but in Kenya 2004, phone spending is big compared to gas+electricity. *breakd sp_house sp_housi sp_rent sp_energ sp_utili sp_el_wa sp_gas sp_phone sp_repai by survey /cells mean. *desc sp_housi sp_utili. /* Some missing data. if (INT_year<>2000 & sp_housi>-1) sp_HOUSE = sp_housi + sp_ENERG. if (INT_year<>2000 & sp_housi>-1 & sp_utili>-1) sp_HOUSE = sp_housi + sp_utili. * If we know both sp_utili and sp_energ, use sp_utili: it's usually bigger than sp_energ. * In the above line, could assume sp_utili is (approximately) zero if sp_utili is missing: it would slightly increase the sample size. if (survey='2002 India' or survey='2007 India' or survey='2012 India' or survey='2017 India') sp_HOUSE = sp_rent + sp_utili. *if (survey='2002 India' or survey='2007 India') sp_housi = sp_rent. if (sp_repai>0) sp_HOUSE = sp_HOUSE + sp_repai. /* sp_repai is only in WAS Egypt, so far. * Define SP_TOTAL and subtotals SP_FOOD, SP_ENERG in each WAS survey separately: SP_TOTAL is usually the sum of each individual spending category in that * survey, but sometimes (e.g. Egypt 2005-6) there's missing data - then, I use the total spending calculated by the organisation collecting the data (to avoid missing data). * In some surveys (e.g. Nigeria), SP_ENERG includes sp_water; if possible, exclude sp_water from SP_ENERG. *compute sp_nFOOD = sp_TOTAL - sp_FOOD. *if (sp_total-9) sp_hHelp = sp_maid. recode sp_hHelp ( -9 = sysmis). * In 2008 Chad & 2009 Cameroon at least, sp_food is sometimes more than sp_total; both surveys asked respondents sp_TOTAL. * If sp_FOOD+sp_HOUSEsp_TOTAL, assume respondent (or interviewer) added up items wrongly. if (sp_FOOD + sp_HOUSE >sp_TOTAL) sp_TOTAL = sp_FOOD + sp_HOUSE. if (sp_FOOD + sp_HOUSE + sp_hHelp >sp_TOTAL) sp_TOTAL = sp_FOOD + sp_HOUSE + sp_hHelp. /* As previous line; sp_hHelp has some missing data. *compute sp_sum = sp_FOOD + sp_HOUSE + sp_hHelp. *corr sp_sum sp_TOTAL. * This list is in the same order as the (/ PPP_now) list near the end of 'combine.sps'. var label sp_cerea 'HH spending (per month): cereals eg rice, corn, wheat flour'. var label sp_cereB 'HH spending (per month): cereals eg rice, flour; bread'. var label sp_bread 'HH spending (per month): bread (& jam, in India)'. var label sp_veget 'HH spending (per month): vegetables/fruit'. var label sp_meatF 'HH spending (per month): meat and fish'. var label sp_fish 'HH spending (per month): fish'. var label sp_meat 'HH spending (per month): meat'. var label sp_oil 'HH spending (per month): cooking oil (& butter/margerine)'. var label sp_oilBr 'HH spending (per month): cooking oil, bread, jam, etc'. var label sp_dairy 'HH spending (per month): milk, dairy products, milk food drinks'. var label sp_br_mi 'HH spending (per month): bread & milk'. var label sp_chees 'HH spending (per month): cheese & dairy products except milk'. var label sp_BDair 'HH spending (per month): bread & dairy products'. var label sp_tuber 'HH spending (per month): tubers eg cassava, sweet potato, potato'. var label sp_grVeg 'HH spending (per month): green vegetables eg spinach'. var label sp_pulse 'HH spending (per month): beans/pulses/nuts, e.g. mung beans, tofu'. var label sp_fruit 'HH spending (per month): fruit'. var label sp_spice 'HH spending (per month): spices eg salt'. var label sp_insta 'HH spending (per month): instant foods (e.g. frozen, tinned)'. var label sp_Fmeal 'HH spending (per month): restaurant meals'. var label sp_mSnak 'HH spending (per month): restaurants & snacks'. var label sp_snack 'HH spending (per month): snacks from street-stalls'. var label sp_biscu 'HH spending (per month): biscuits, etc'. var label sp_ices 'HH spending (per month): ice-creams'. var label sp_cofee 'HH spending (per month): coffee, tea, etc'. var label sp_other 'HH spending (per month): other provisions'. var label sp_FOOD 'HH spending (per month): ALL FOOD excluding restauraunts'. var label sp_gas 'HH spending (per month): gas/kerosene'. var label sp_elect 'HH spending (per month): electricity'. var label sp_el_wa 'HH spending (per month): electricity + water'. var label sp_ENERG 'HH spending (per month): gas+electricity'. /* Often the same as sp_utili. var label sp_phone 'HH spending (per month): telephone bill (including mobile)'. var label sp_utili 'HH spending (per month): household utilities: electricity, gas, water, phone, etc'. var label sp_rent 'HH spending (per month): house rent'. var label sp_mortg 'HH spending (per month): mortgage payment'. var label sp_housi 'HH spending (per month): house rent or mortgage'. /* Often the same as sp_rent. var label sp_loanR 'HH spending (per month): loan repayment'. /* mortgage payment is with rent, not in sp_loanR. var label sp_repai 'HH spending (per month): house repairs'. var label sp_HOUSE 'HH spending (per month): RENT + GAS/ELECTRICITY/WATER'. var label sp_tobac 'HH spending (per month): tobacco'. var label sp_drink 'HH spending (per month): alcoholic drinks'. var label sp_CigDr 'HH spending (per month): tobacco + drink'. var label sp_gambl 'HH spending (per month): gambling, lottery'. var label sp_tFuel 'HH spending (per month): fuel for car (& insurance)'. var label sp_trans 'HH spending (per month): transport'. var label sp_cloth 'HH spending (per month): clothes/footwear inc repairs'. var label sp_clean 'HH spending (per month): cleaning products, eg detergent'. var label sp_toile 'HH spending (per month): personal toiletries'. var label sp_laund 'HH spending (per month): laundry services'. var label sp_maid 'HH spending (per month): servant/maid'. var label sp_cCare 'HH spending (per month): child-care'. var label sp_hHelp 'HH spending (per month): laundry, maid (& child-care)'. var label sp_medic 'HH spending (per month): paid to hospital/doctor, medicine, etc'. var label sp_leisu 'HH spending (per month): recreation/entertainment'. var label sp_leisT 'HH spending (per month): recreation/entertainment + tobacco/alcohol'. var label sp_haird 'HH spending (per month): make-up & hairdressers'. var label sp_paper 'HH spending (per month): newspapers & magazines'. var label sp_cerem 'HH spending (per month): ceremonies/rituals, e.g. wedding'. var label sp_chari 'HH spending (per month): gifts to charities'. var label sp_holid 'HH spending (per month): travel/vacations (holidays)'. var label sp_furni 'HH spending (per month): furniture e.g. beds, sheets, towels'. var label sp_remit 'HH spending (per month): regular non-food gifts outside HH'. var label sp_educ 'HH spending (per month): education: school or college fees'. var label sp_lessn 'HH spending (per month): private lessons'. var label sp_tax 'HH spending (per month): tax, including property & vehicle'. var label sp_TOTAL 'Total HH spending (per month)'. *. * Individual spending (only in WAS surveys after 2003, except sp_Hmeal & sp_Wmeal - which are only in bm92): . var label sp_Hmeal "Spending by husband/male respondent (per month): restaurant meals". var label sp_Wmeal "Spending by wife/female respondent (per month): restaurant meals". var label sp_Hdrnk "Spending by husband/male respondent (per month): alcohol". var label sp_Wdrnk "Spending by wife/female respondent (per month): alcohol". var label sp_HClth "Spending by husband/male respondent (per month): clothes & shoes". var label sp_WClth "Spending by wife/female respondent (per month): clothes & shoes". var label sp_HGamb "Spending by husband/male respondent (per month): lottery/gambling". var label sp_WGamb "Spending by wife/female respondent (per month): lottery/gambling". var label sp_Hsoft "Spending by husband/male respondent (per month): soft drink". var label sp_Wsoft "Spending by wife/female respondent (per month): soft drink". var label sp_Htoba "Spending by husband/male respondent (per month): tobacco etc". var label sp_Wtoba "Spending by wife/female respondent (per month): tobacco etc". var label sp_Hcafe "Spending by husband/male respondent (per month): in cafes". /* Egypt: "smoking (cigarettes, sheesha,cofeeshop etc)". var label sp_Wcafe "Spending by wife/female respondent (per month): in cafes". /* Egypt: "smoking (cigarettes, sheesha,cofeeshop etc)". var label sp_Hpers "Spending by husband/male respondent (per month): care e.g. haircut". var label sp_Wpers "Spending by wife/female respondent (per month): care e.g. haircut". var label sp_Hmobi "Spending by husband/male respondent (per month): mobile phone". var label sp_Wmobi "Spending by wife/female respondent (per month): mobile phone". var label sp_Hweb "Spending by husband/male respondent (per month): internet". var label sp_Wweb "Spending by wife/female respondent (per month): internet". var label sp_Hremt "Spending by husband/male respondent (per month): to family outside hh". var label sp_Wremt "Spending by wife/female respondent (per month): to family outside hh". var label sp_Hchar "Spending by husband/male respondent (per month): donations e.g. church". var label sp_Wchar "Spending by wife/female respondent (per month): donations e.g. church". var label sp_Hcars "Spending by husband/male respondent (per month): transport equipment e.g. car". /* Egypt only. var label sp_Wcars "Spending by wife/female respondent (per month): transport equipment e.g. car". /* Egypt only. var label sp_Hleis "Spending by husband/male respondent (per month): entertainment". /* Egypt only. var label sp_Wleis "Spending by wife/female respondent (per month): entertainment". /* Egypt only. var label sp_HOthr "Spending by husband/male respondent (per month): other". /* Egypt only. var label sp_WOthr "Spending by wife/female respondent (per month): other". /* Egypt only. var label SAVING 'Household saving (per month)'. *. var label earn_hus "Earnings of husband after tax, per month". var label earn_wif "Earnings of wife after tax, per month". var label earnResp "Earnings of respondent after tax, per month". var label earnSpou "Earnings of respondent's spouse after tax, per month". var label earnOthr "Earnings of other household member(s) after tax, per month". var label HHincome "Total household income, per month". var label HHincEQI "Household equivalent income, per month: income, adjusted for number of people in household". if (int_year<>1994 & sp_maid=0) hireMaid=0. if (int_year<>1994 & sp_maid>0) hireMaid=1. var label hireMaid 'Does the HH hire a full-time maid?'. val label hireMaid 0"don't hire a fulltime maid" 1"hire full-time maid" 2"2+ full-time maids". var label skipMeal 'Frequency of cutting meals in the last year'. val labels skipMeal 0 'never' 1 'only 1 or 2 months' 2 'some months, but not every month' 3 'almost every month'. *@ DURABLES OWNERSHIP. *breakd car by survey. recode TVcolour,TVbw (sysmis=-9). if (INT_year<>1994 & INT_year<>1997) TVcolour = -6. if (INT_year<>1994 & INT_year<>1997) TVbw = -6. if (TVbw = 0 or TVcolour = 0) TV = 0. /* For 1994 & 1997 only. if (TVbw > 0 or TVcolour > 0) TV = 1. /* For 1994 & 1997 only. if ( survey='2002 India' & sp_phone=0) phone = 0. if ( survey='2002 India' & sp_phone>0) phone = 1. /* Could be a mobile phone. * Livestok is new to Egypt 2005. if (cows=0 & goats=0) livestok = 0. if (cows=1) livestok = 1. if ( goats=1) livestok = 1. var label CAR 'Does the HH own a car?'. var label motorcyc 'Does the HH own a motorcycle/moped?'. var label bicycle 'Does the HH own a bicycle?'. var label TV 'Does the HH own a television?'. var label TVcolour 'Does the HH own a colour television?'. var label TVbw 'Does the HH own a black-and-white television?'. var label TVbox 'Does the HH own a set-top box, to use digital TV on a normal TV set?'. var label SATELITE 'Does the HH have satelite television (or cable TV)?'. var label VIDEO 'Does the HH own a video-cassette player?'. var label Hi_Fi 'Does the HH own a music system (hi-fi)?'. var label RADIO 'Does the HH own a radio?'. var label COMPUTER 'Does the HH own a computer?'. var label use_PC 'Do you use a computer (including a laptop)?'. var label e_mail 'Do you have e-mail address?'. var label INTERNET 'Does the HH have access to the internet on a home PC?'. var label PHONE 'Is there a (fixed line) telephone in this house/flat?'. var label MOBILE 'Does the HH own a mobile (cellular) phone?'. var label VACUUM 'Does the household own a vacuum-cleaner?'. var label WASHMACH 'Does the HH own a washing-machine?'. var label FRIDGE 'Does the HH own a refrigerator (or fridge/freezer)?'. var label FOODPROC 'Does the HH own an electric foodprocessor?'. var label WETGRIND 'Does the HH own an electric food grinder?'. var label cookType 'Type(s) of cooker the household owns'. var label presCook 'Does the HH own a pressure-cooker?'. var label riceCook 'Does the HH own a rice cooker?'. var label TOASTER 'Does the HH own an electric toaster?'. var label purifier 'Does the HH own a water purifier/Reverse Osmosis system?'. var label MICROWVE 'Does the HH own a microwave oven?'. var label fan 'Does the HH own an electric fan?'. var label ELECTRIC 'Is the home connected to electricity?'. var label generatr 'Does the HH own a electricity generator (for power cuts)?'. var label WATERtap 'Water source for the household'. var label WATERhot 'Is there hot water on tap in the home?'. var label TOILET 'Is there a flush toilet/bathroom in the home?'. var label cows 'Does the HH own any cows?'. var label goats 'Does the HH own any goats?'. var label farmland 'Does the HH own any farmland?'. var label clock 'Household owns: clock or watch'. var label bed 'Household owns: bed'. var label videoDVD 'Household owns: video and/or DVD'. var label sewMachi 'Household owns: sewing machine'. var label air_cond 'Household owns: air-conditioner'. var label purifier 'Household owns: water purifier/Reverse Osmosis system'. var label truck 'Household owns: truck, taxi, van, or bus'. var label livestok 'Household owns: farm livestock'. var label farmEqip 'Household owns: farm equipment'. var label secondHm 'Household owns: commercial/residential building: not own home'. val label TV,TVcolour,TVbw,TVbox,satelite,video,Hi_Fi,radio,computer,e_mail,internet,mobile, vacuum,washMach,fridge,foodProc,wetGrind,presCook,riceCook,toaster,microwve, toilet generatr motorcyc bicycle fan cows goats farmland 0'do not own' 1'own' 2'own 2' 3'own 3' 4'own 4' 5'own 5' 6'own 6+'. val label car 0'do not own' 1'own' 2'own 2' 3'own 3' 4'own 4' 5'own 5' 6'own 6' 7'own 7' 8'own 8'. val label phone 0'no' 1'have a phone' 2 'have 2 phones' -4'yes, but it is broken'. val label use_PC 0'no' 1'yes'. val label cookType lighting 0'none' 1'electric' 2'gas(/LPG)' 3'kerosene(/paraffin)' 4'firewood' 5'charcoal' 6'coal/lignite' 7'dung' 9'solar' 11'electric & gas' 12'electric & kerosene' 13'gas & kerosene' 14'electric & gas & kerosene' 99'cooker (unspecified)' 1.5'gas or electric' 2.5'electric or gas, and kerosene'. /* These labels are for Nigeria: cooker may be gas OR electric. * For Kenya 2004, I use cookType value labels for lighting also. var label Lighting 'Main source of lighting'. val label ELECTRIC 0'not connected' 1'connected' 2'connected but not switched on'. val label WATERhot 0 'do not have a hot water tap' 1'hot water tap(s) in home'. val label WATERtap 0 'do not have a water tap' 10 'from stream/river/canal' 15 'from spring' 20 'rain water' 30 'other: unspecified' 40 'open public well: no pump' 41 'open public well (dk if pump)' 42 'open public well: pump' 45 'open private well: no pump' 46 'open private well (dk if pump)' 47 'open private well: pump' 50 'protected public well: no pump' 51 'protected public well (dk if pump)' 52 'protected public well: pump' 55 'protected private well: no pump' 56 'protected private well (dk if pump)' 57 'protected private well: pump' 70 'buy water' 80 'bottled mineral water' 90 'tap in the street' 95 'tap in the yard/plot' 100 'tap in home'. val label TOILET 0'no' 0.5'toilet outside, e.g. pit' 0.8'traditional bucket flush toilet' 0.9'traditional tank flush toilet' 1'toilet in the house' 2'have 2 toilets' 3'have 3 toilets' 4'have 4 toilets' 5'have 5 toilets' 6'have 6+ toilets'. *@ TYPE OF ACCOMMODATION. var label homeType 'is respondent a slum/street-dweller?'. val label homeType val label homeType 0 "street-dweller" 10 "self-built shack: temporary (kuchcha jhuggi)" 20 "self-built shack: permanent (pucca jhuggi)" 25 "self-built shack in a backyard" 30 "home surrounded by animal waste/puddles/refuse/stable" 40 "cortico (tenement) dweller" 43 "garage/ modified garage/ rooms in the back" 45 "traditional hut" 46 "Rondavel/ Zozo hut" 50 "other: unspecified" 55 "caravan or mobile home" 56 "single room" 58 "two rooms" 59 "simple wooden house" 60 "roomed house (RDP house), or room+parlour" 61 "matchbox type house or 51/9 (3-4 rooms): separate stand/yard" 63 "improved matchbox house on separate stand/yard, or mini-flat" 70 "simple cement house" 80 "part of a house/ share a house" 84 "granny flat on this property/ flatlet" 85 "second house/ cottage on this property" 90 "a unit in a block of flats" 97 "semi-detached or joint house" 98 "townhouse or cluster house in complex" 99 "suburban type house (2 or more bedrooms, inside bathroom)" 100 "OK house/apartment". * High score is OK house: my estimate. Could be systematic, eg: permanent? waterproof? water supply?. * Hard to match to Indonesia data - which has many variables. * India 2002 qnaire has 'self-built home'=1: I meant a normal home, but it could be interpreted as shack. * Could use phone to identify if home is permanent? - but some phones are mobile/cellphones. * I created a new system to classify building materials, where a higher score means less permanent. 'Masonry' is a vague term. value label homeFlor homeWall homeRoof 0'carpet/vinyl/parquet or polished wood' 1'concrete' 2'ceramic/marble/granite/stone' 3'(red) brick & cement' 4'masonry (concrete/stone/etc)' 5'asbestos cement' 6'cement/tiles/terazzo' 7'roof tiles' 8'zinc sheets' 9'iron sheets' 10'hardwood/board/plywood' 11'bamboo/plaited bamboo slats' 12'grass thatch' 13'foliage/palm leaves' 14'soil/dirt/mud' -4'other'. * Make an aproximation for surveys in which homeWall & homeRoof aren't known. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=10) homeWall = 10. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=20 or homeType=25) homeWall = 9. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=45 or homeType=46) homeWall = 14. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=60 or homeType=61 or homeType=63) homeWall = 3. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=80 or homeType=84 or homeType=85) homeWall = 3. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=40 or homeType=90) homeWall = 1. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=97 or homeType=98 or homeType=99 or homeType=100) homeWall = 3. * In S.Africa, I think 'matchbox house' & 'RDP house' both have brick walls & tin roof. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=10) homeRoof = 13. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=20 or homeType=25) homeRoof = 9. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=45 or homeType=46) homeRoof = 12. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=60 or homeType=61 or homeType=63) homeRoof = 9. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=80 or homeType=84 or homeType=85) homeRoof = 7. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=40 or homeType=90) homeRoof = 4. if (int_year<2001 or live_cit<7 or int_year=2003 or int_year=2005) & (homeType=97 or homeType=98 or homeType=99 or homeType=100) homeRoof = 7. * Could also estimate homeFlor, using homeType. var lab homeWall 'Building materials used to build home: walls'. var lab homeRoof 'Building materials used to build home: roof'. var lab homeFlor 'Building materials used to build home: floor'. var labels tenure 'Home is owned or rented'. val labels tenure 100'self owned' 90'owned jointly' 80'owned by family member' 70'buying on mortgage' 50'other: unspecified' 30'rented (contracted)' 20'not self-owned, but no rent paid'. Document 'Work, Attitudes & Spending in India, Brazil, South Africa, Indonesia, Nigeria, Kenya & Egypt': This SPSS dataset combines data from the following household surveys: 1992 survey: Bombay and Madras urban India (IMRB); 1994 survey: Rio de Janeiro & Sao Paulo urban Brazil (Marplan); 1997 survey: Bombay, Madras, Delhi, Calcutta urban India (IMRB); 2000 survey: eleven cities urban South Africa (Markinor) 2001 survey: Palembang,Jakarta,Bandung,Surabaya urban Indonesia (Univ.Indonesia); 2002 survey: Palembang,Jakarta,Bandung,Surabaya urban Indonesia (Univ.Indonesia); 2002 survey: Bombay,Madras,Delhi,Calcutta,Kerala,Patna urban India (IMRB); 2003 survey: 37 locations urban+rural Nigeria (RMS); 2004 survey: 44 locations urban+rural Kenya (SBO); 2005 survey: 37 locations urban+rural Nigeria (RMS); 2005 survey: seven governorates urban+rural Egypt (Cairo University); 2007 survey: 11 cities urban India (IMRB); 2008 survey: some but not all regions (some unsafe, due to insurgents) urban+rural Chad (Cible); 2009 survey: all regions urban+rural Cameroon (Cible); 2011 survey: several regions urban+rural Congo-Brazzaville (Cible); 2012 survey: 11 cities urban India (IMRB) 2017 survey: 11 cities urban India (IMRB) All were commissioned by John Simister, Economics & International Business Dept, Manchester Metropolitan University, UK; Fieldwork was carried out by the Indian Market Research Bureau Ltd; Marplan Brasil ltda.; Markinor Pty Ltd.; University of Indonesia; Research & Marketing Services Ltd; SBO Research Ltd (formerly called Strategic Business Opportunities Ltd); CSSA: FEPS, University of Cairo; Cible Ltd. The data was thoroughly checked by John Simister, and some new variables calculated, before being deposited (with machine-readable documentation) at the ESRC Data Archive (University of Essex, UK). See the SPSS syntax files for more information on variables included in the dataset, and how derived variables were created. & + The 1992 data was deposited earlier at the UK Data Archive (study SN:3290); I call it 'bm92'. For each variable in this dataset, I feel that variable is comparable between different surveys, except the following differences between bm92 & this dataset: * In bm92, time-use variables COOK_HUS, COOK_WIF, WASH_HUS, & WASH_WIF were in hours PER DAY (in this dataset, time-use is in hours per week); * In bm92, country-of-birth codes (for people born outside India) were not internet country codes (which this survey uses). & + There are important differences between the sample frame for these surveys; hence comparisons between countries should be used with caution. The BM92 and RS94 surveys were based on a random survey of households, so each sample is representative of the cities studied. However, the BMDC97 sample was limited to married couples (to increase the number of couples: WAS surveys were intended to study wives control over household spending); SA2000 includes 2,000 urban and 1,500 rural households, but most questions were not asked of the rural subsample. & + The India (1997) survey was written in English, and (where necessary) IMRB interviewers translated each question to language of the respondent - shown in variable INT_LANG. The Brazil (1994) questionnaire was translated from English to Portuguese, by Marplan Brasil; all interviews were carried out in Portuguese. Documentation files for this dataset include the Portuguese version of the questionnaire (as used in the actual survey, but retyped by John Simister), as well as a translation from Portuguese back into English. The The SA2000 survey was written in English and translated by the interviewer where necessary. The 2001 Indonesian survey is in Bahasa Indonesian. & + I am grateful to the employees at the organisations which carried out fieldwork; their advice on improving the questionnaire was very helpful - I am especially grateful to Virginia Silva e Silva (Marplan Brasil); Anneke Greyling (Markinor); Wiyono Nur Hadi & Hendro Hendratno (University of Indonesia); Mariam Fagbemi (RMS); Peninnah Mukiri (SBO); and Fatma El Zanaty (Cairo University). & John Simister, March 2017. *END DOCUMENT.