Date post: | 20-Dec-2015 |
Category: |
Documents |
View: | 216 times |
Download: | 0 times |
www.ipums.org/international 1
Calibrating census microdata Calibrating census microdata against a gold standard against a gold standard (employment survey): women (employment survey): women in the workforce, in the workforce, Mexico 1970, 1990 and 2000Mexico 1970, 1990 and 2000 * * * * * *Robert McCaa, Rodolfo Gutierrez and Robert McCaa, Rodolfo Gutierrez and Gabriela Vasquez, Minnesota Population Gabriela Vasquez, Minnesota Population Center Center www.ipums.org/internationalwww.ipums.org/internationalCalibrate, v. 1864. a. Calibrate, v. 1864. a. trans.trans.
...to graduate a gauge of any kind ...to graduate a gauge of any kind with allowance for its irregularities.with allowance for its irregularities.
The Oxford English Dictionary OnlineThe Oxford English Dictionary Online (Oxford: Clarendon Press, 2001)(Oxford: Clarendon Press, 2001)
www.ipums.org/international 2
IPUMSIPUMSi i
Integrated Public Use Series Integrated Public Use Series InternationalInternational
Minnesota Population CenterMinnesota Population CenterUniversity of MinnesotaUniversity of MinnesotaPrincipal investigators:Principal investigators:
historians:historians: Steven Ruggles, director MPC Steven Ruggles, director MPC
Robert McCaa, Matt SobekRobert McCaa, Matt Sobekdemographers:demographers:
Deborah Levison, Miriam King Deborah Levison, Miriam Kingwww.ipums.org/internationalwww.ipums.org/international
www.ipums.org/international 3
IPUMSIPUMSii goals goals » 1. 1. InventoryInventory the world’s census microdata the world’s census microdata
a.a. historicalhistoricalb. contemporaryb. contemporary
» 2. 2. PreservePreserve endangered microdata and metadata endangered microdata and metadataa. a. contract preservation with repositoriescontract preservation with repositoriesb. b. archive validated copiesarchive validated copies
* * ** * *» 3. 3. IntegrateIntegrate census microdata and metadata census microdata and metadata
of selected countries on all continentsof selected countries on all continentsusing UN, ECE, and other standardsusing UN, ECE, and other standards
» 4. 4. DisseminateDisseminate resulting database without charge resulting database without charge
with full access to all who agree to non-with full access to all who agree to non-disclosuredisclosure
www.ipums.org/international 4
IPUMSIPUMSii integration principles integration principles
» 1. 1. RespectRespect absolute anonymity absolute anonymity» 2. 2. PreservePreserve all original data, except adjustments all original data, except adjustments
to insure confidentiality (top codes blurrings, to insure confidentiality (top codes blurrings, masking, re-ordering, etc.)masking, re-ordering, etc.)
» 3. 3. HarmonizeHarmonize (integrate) codes for countries (integrate) codes for countries using international standards.using international standards.
occupation: occupation: ISCO (detailed, general)ISCO (detailed, general)education: education: ISCED “ “ISCED “ “family: family: IPUMS “ “IPUMS “ “
» 4. 4. EnhanceEnhance with constructed variables with constructed variables
» 5. 5. CalibrateCalibrate microdata against “gold standards” microdata against “gold standards”
www.ipums.org/international 5
IINNTTEEGGRRAATTEESS
II
PP
UU
MM
SSii
12 projects 12 projects startedstarted
USAUSA 1850-1880, 1900-1920, 1850-1880, 1900-1920, 1940-20001940-2000
ColombiaColombia 1964, 1973, 1985, 1993 1964, 1973, 1985, 1993FranceFrance 1962, 1968, 1975, 1982, 1962, 1968, 1975, 1982, 19901990MexicoMexico 1960, 1970, 1990, 20001960, 1970, 1990, 2000VietnamVietnam 1989, 19991989, 1999KenyaKenya 19691969??, 1979, 1979??, 1989, 1999, 1989, 1999East-West CenterEast-West Center (Hawaii) (Hawaii)Centro Latinoamericano de Centro Latinoamericano de DemografiaDemografiaBrazilBrazil 1960, 1970, 1980, 1991, 1960, 1970, 1980, 1991, 20012001SpainSpain 1981, 1991, 20011981, 1991, 2001 HungaryHungary 1980, 1990, 20011980, 1990, 2001ChinaChina?? 1982, 1990, 20001982, 1990, 2000
www.ipums.org/international 6
INTEGRATES
Photos of the IPUMS-Colombia teamFebruary-March, 2000:
5 statisticians from DANE +8 academics (3 universities)
IPUMSi
Standard:UN/Eurostat Standard:UN/Eurostat Principles & Recs...Principles & Recs...
Census documentation Census documentation compiled for the compiled for the Colombia projectColombia project
www.ipums.org/international 7
DDIISSSSEEMMIINNAATTEESS
II
PP
UU
MM
SSii
End-User license agreement End-User license agreement
protects confidentialityprotects confidentiality
assures proper use assures proper use
User selects User selects
countries, countries,
censuses,censuses,
cases, cases,
variables, variables,
samples--assists comparative research samples--assists comparative research
Open architecture softwareOpen architecture software
International web-based International web-based access systemaccess system
www.ipums.org/international 8
Calibration testCalibration testMexico 1990, 2000: FLFPMexico 1990, 2000: FLFP
census microdata census microdata vs. employment survey vs. employment survey
Employment surveys: Employment surveys: date from the 80s; many probing questions date from the 80s; many probing questions finely tuned instrument collected by trained finely tuned instrument collected by trained interviewersinterviewersurban--lack national coverageurban--lack national coverage
Census microdata: Census microdata: strength: national coverage back to 1960 strength: national coverage back to 1960 weakness: untrained interviewers, one questionweakness: untrained interviewers, one questionomits many working women, particularly informal omits many working women, particularly informal workersworkers
Purpose of paper: Purpose of paper: calibrate census microdata w/ employment calibrate census microdata w/ employment
surveyssurveys
www.ipums.org/international 9
Table 1. Selected microdata Table 1. Selected microdata samples of Mexico, 1960 - 2000 samples of Mexico, 1960 - 2000
YearYear Type Type Sample SizeSample SizeDensity (%)Density (%)
19601960** Census Census 502,702502,702 1.51.5
19701970** Census Census 480,265480,265 1.01.0
19801980 Census Census No sample available due to No sample available due to earthquake damage earthquake damage
19901990** Census Census 802,774802,774 1.01.0
1990, ENEU (survey)1990, ENEU (survey) 172,233172,233 0.20.2
20002000** Census Census 10,099,182 10,099,182 10.0 10.0
20002000 ENEU (survey) ENEU (survey) 562,471562,471 0.60.6
((**to be integrated in IPUMSto be integrated in IPUMSii))
www.ipums.org/international 10
The problem (table 2)The problem (table 2)Mexico’s “global” female labor force Mexico’s “global” female labor force
participation rate (12-64 years)participation rate (12-64 years)
microdatamicrodata 1990199020002000
survey (ENEU):survey (ENEU): 34.6% 34.6% 43.3%43.3%
national census:national census: 20.6% 20.6% 32.9%32.9%
14.014.0 10.410.4A better comparisonA better comparisonControl for survey (ENEU) sampling Control for survey (ENEU) sampling
frame: frame: 16 cities in 199016 cities in 1990
survey (16 cities): 34.6%survey (16 cities): 34.6% 41.7%41.7%
census (16 cities):census (16 cities): 29.0% 29.0% 40.2%*40.2%*
5.65.6 1.5 1.5* includes responses from LFP questions 1 (activity) & 2 * includes responses from LFP questions 1 (activity) & 2 (verification).(verification).
www.ipums.org/international 11
2000 census--two questions on 2000 census--two questions on LFP: LFP:
1: “Last week did (NAME)...” 1: “Last week did (NAME)...” ...”Question 1:Question 1:
Last week (Name):Last week (Name):
Did you work? Did you work? 27.5%27.5%
Had work? Had work? 0.40.4
Look for work? Look for work? 0.30.3
Are you a Are you a student?student?
housewife?housewife?
retired?retired?
permanently permanently incapacitated?incapacitated?
Did you not work?Did you not work?
2000 census--two questions on 2000 census--two questions on LFP: LFP:
1: “Last week, did (NAME)...?” 1: “Last week, did (NAME)...?” 2: “Besides (...), did (NAME)...?”2: “Besides (...), did (NAME)...?”Question 1:Question 1:
Last week (Name):Last week (Name):
Did you work? Did you work? 27.5%27.5%
Had work? Had work? 0.40.4
Looked? Looked? 0.30.3Q. 1&2: combined Q. 1&2: combined student/wrkd student/wrkd 0.50.5 housewife/wr housewife/wr 3.73.7 retired/wrkd retired/wrkd 0.00.0 other/w? other/w? 0.40.4 no reply/w? no reply/w? 0.00.0
Question 2:Question 2:
Did you help in a Did you help in a family family business?business?
Sell some Sell some product?product?
Make some Make some product to sell?product to sell?
Help on a farm or Help on a farm or with livestock?with livestock?
Or in exchange Or in exchange for pay did you for pay did you do some other do some other activity?activity?
www.ipums.org/international 12
StructureStructure FLFP RatesFLFP Rates
SurveySurvey CensusCensus SrvySrvy CnssCnss
TotalTotal 62,248 62,248 63,92963,929 34.634.6 29.029.0
EducationEducation
Less than 6 yearsLess than 6 years 20.920.9 21.7 21.7 29.329.3
20.120.1
Completed primaryCompleted primary 34.734.7 34.8 34.8 27.627.6 21.121.1
Completed middleCompleted middle 20.420.4 24.3 24.3 31.331.3
37.937.9
Post-middle (10+)Post-middle (10+) 23.923.9 19.3 19.3 53.153.1
42.242.2
Marital StatusMarital Status
Married (all forms)Married (all forms) 48.248.2 50.4 50.4 27.727.7 21.321.3
Not in unionNot in union 51.851.8 49.6 49.6 41.441.4
36.936.9
Table 3. Urban Females, 1990Table 3. Urban Females, 1990(aged 12-64)(aged 12-64)
www.ipums.org/international 13
StructureStructure FLFP RatesFLFP Rates
SurveySurvey CensusCensus SrvySrvy CnssCnss
TotalTotal 124,051 1,073,222 124,051 1,073,222 41.741.7 40.240.2
EducationEducation
Less than 6 yearsLess than 6 years 14.914.9 15.9 15.9 35.335.3
31.031.0
Completed primaryCompleted primary 30.830.8 28.7 28.7 32.132.1 30.130.1
Completed middleCompleted middle 28.528.5 19.7 19.7 47.047.0
41.041.0
Post-middle (10+)Post-middle (10+) 25.825.8 36.0 36.0 51.151.1
51.951.9
Marital StatusMarital Status
Married (all forms)Married (all forms) 51.951.9 52.9 52.9 35.135.1 34.034.0
Not in unionNot in union 48.148.1 47.1 47.1 48.948.9
47.347.3
Table X. Females 2000: Urban Table X. Females 2000: Urban (same 16 cities as in ENEU 1990) (same 16 cities as in ENEU 1990)
(table not in paper)(table not in paper)
www.ipums.org/international 14
Female labor force participationFemale labor force participationENEU (indicator) vs. CensusENEU (indicator) vs. Census
Variable B S.E. Wald df Sig R Exp(B)
1990: -.2212 .0135 267.2291 1 .0000 -.0412 .8015
2000: .0860 .0067 163.1781 1 .0000 .0100 1.0898Model source effect taking into account age, marital status and education.Model source effect taking into account age, marital status and education.
Table Y. Logistic Regression: Table Y. Logistic Regression: Source Source (Females 1990, 2000; same 16 cities as (Females 1990, 2000; same 16 cities as
in ENEU 1990)in ENEU 1990) (table not in paper)(table not in paper)
InterpretationInterpretationIf for both sources weights are correct and slight If for both sources weights are correct and slight structural differences are taken into account:structural differences are taken into account:
1990 census under-reported 20% of FLFP.1990 census under-reported 20% of FLFP.
2000 census over-reports FLFP by 9%.2000 census over-reports FLFP by 9%.
www.ipums.org/international 15
LFP by sex and marital statusLFP by sex and marital statusMexico 1990 and 2000 (national figures)Mexico 1990 and 2000 (national figures)
The iron grip of m arriage on working for pay is w eakeningMarriage dam pens partic ipation for fem ales in c ontras t to m ales
F ig . 1 . D e ta ile d lab o r fo rc e ra te s s ta tus : M e x ic o , 1 9 9 0 , 2 0 0 0Married inc ludes religious or consensual as w ell as legal unions
R a t e s o f m a r ried w o m e n w e re les s th a n h a lf th o s e o f s in g le , w id ow e d , s e p a ra t e d o r d iv o rc e d
pe
rce
nt
F em ales : 1990age
12 15 20 25 30 35 40 45 50 55 60 65
0
20
40
60
80
100
Sing le
M ar r ied
W idSe pD v
T h e g ap b e twe e n m a rr ie d w o m en a n d o th e rs s h ra n k , bu t t h e d if fe re n c e re m a in s s u b s ta n t ia l
F em ales : 2000age
12 15 20 25 30 35 40 45 50 55 60 65
0
20
40
60
80
100
Sing le
M ar r ied
W idSe pD v
R a te s o f m a rr ie d m a le s a re m a rk e d ly h ig he r t h a n f o r o t h e rs
pe
rce
nt
M ales : 1990age
12 15 20 25 30 35 40 45 50 55 60 65
0
20
40
60
80
100
Sing le
M ar r iedW idSe pD v
T h e g a p n a r ro w e d s l ig h t ly ,p a r t ic u la rly f o r t h e w id o w e d , s e p a ra te d a nd d iv o rc e d
Males : 2000age
12 15 20 25 30 35 40 45 50 55 60 65
0
20
40
60
80
100
Sing le
M ar r iedW idSe pD v
FemaleFemaless
MalesMales19901990 20002000
www.ipums.org/international 16
Bigges t change is for w om en w ith les s than 7 years of s choolingLittle = < 6 years ; P rim ary = 6-8; Middle = 9; Higher = 9+ years
Fig. 2. Schooling and Marriage strongly influence female work ratesMarried inc ludes all form s of unions
F ew wit h les s t han 7 y ea rs o f s c hooling wo rk ed fo r pay
pe
rce
nt
1990: Not C urrently Marr iedage
12 15 20 25 30 35 40 45 50 55 60 65
0
20
40
60
80
100
litt le
prim ary
m idd le
h igher
G rea t es t inc reas es a re fo r t hos e wit h les s than 7 y ea rs o f s c hoo ling
2000: Not C urrently Marr iedage
12 15 20 25 30 35 40 45 50 55 60 65
0
20
40
60
80
100
litt le
prim arym idd le
h igher
pe
rce
nt
1990: Marr iedage
12 15 20 25 30 35 40 45 50 55 60 65
0
20
40
60
80
100
litt leprim ary
m idd le
h igher
2000: Marr iedage
12 15 20 25 30 35 40 45 50 55 60 65
0
20
40
60
80
100
litt leprim arym idd le
h igher
Marriage and education strongly Marriage and education strongly affect FLFP affect FLFP (Mexico 1990 and 2000, national (Mexico 1990 and 2000, national
figures)figures)
19901990 20002000MarriedMarried
NotNot
www.ipums.org/international 17
Figure 3. Female labor force participation rates by ageUnited States, 1880-1990 and Mexico, 1970, 1990, 2000
One of the great social transformations of the twentieth centuryMarried women: USA 1880-1990; Mexico 1970, 1990, 2000
pe
rce
nt
Fig. 3a. Labor force participation rate: Married womenedad
16 20 25 30 35 40 45 50 55 60 65
0
20
40
60
80
100
USA1880
USA1910USA1940
USA1970
USA1990
pe
rce
nt
edad
16 20 25 30 35 40 45 50 55 60 65
0
20
40
60
80
100
Mex1970
pe
rce
nt
edad
16 20 25 30 35 40 45 50 55 60 65
0
20
40
60
80
100
Mex1990
pe
rce
nt
edad
16 20 25 30 35 40 45 50 55 60 65
0
20
40
60
80
100
Mex2000
www.ipums.org/international 18
ConclusionsConclusions
Mexican census microdata may be more Mexican census microdata may be more informative, even about FLFP, than researchers informative, even about FLFP, than researchers thinkthink
Mexican census microdata on FLFP display Mexican census microdata on FLFP display remarkable coherence in time and spaceremarkable coherence in time and space
““Chorus of calamity” on Mexican FLFP may Chorus of calamity” on Mexican FLFP may overlook overlook
enormous changes in educationenormous changes in education
weakening power of patriarchy over married womenweakening power of patriarchy over married women
real advances of women in the workforce real advances of women in the workforce
2000 microdata tell the story2000 microdata tell the story
Calibration helps weigh strengths and Calibration helps weigh strengths and weaknesses of sourcesweaknesses of sources
www.ipums.org/international 19
paper is on conference CDpaper is on conference CD
or contact:or contact:[email protected]@umn.edu
* * * * * ** * * * * *Thank youThank you