Calibrating census microdata against a gold standard employment survey: women in the workforce, Mexi


1
Calibrating census microdata against a gold
standard employment survey women in the
workforce Mexico 1970 1990 and 2000
Robert McCaa Rodolfo Gutierrez and
Gabriela Vasquez Minnesota Population Center
www.ipums.org/international
Calibrate v. 1864. a. trans. ...to graduate a
gauge of any kind with allowance for its
irregularities.The Oxford English Dictionary
Online Oxford Clarendon Press 2001
2
Minnesota Population CenterUniversity of
MinnesotaPrincipal investigatorshistorians
Steven Ruggles director MPC Robert McCaa Matt
Sobekdemographers Deborah Levison Miriam
Kingwww.ipums.org/international
IPUMSi Integrated Public Use Series
International
3
IPUMSi goals

• 1. Inventory the worlds census microdata
• a. historicalb. contemporary
• 2. Preserve endangered microdata and metadataa.
  contract preservation with repositoriesb.
  archive validated copies
• 3. Integrate census microdata and metadata
• of selected countries on all continents
• using UN ECE and other standards
• 4. Disseminate resulting database without charge
• with full access to all who agree to
  nondisclosure

4
IPUMSi integration principles

• 1. Respect absolute anonymity
• 2. Preserve all original data except adjustments
  to insure confidentiality top codes blurrings
  masking reordering etc.
• 3. Harmonize integrate codes for countries
  using international standards. occupation
  ISCO detailed general education ISCED
  family IPUMS
• 4. Enhance with constructed variables
• 5. Calibrate microdata against gold standards

5
INTEGRATES
IPUMSi
12 projects started
USA 18501880 19001920 19402000 Colombia
1964 1973 1985 1993 France 1962 1968
1975 1982 1990 Mexico 1960 1970 1990
2000 Vietnam 1989 1999 Kenya 1969? 1979? 1989
1999 EastWest Center Hawaii Centro
Latinoamericano de Demografia
1st Release May 2002 PAA
Brazil 1960 1970 1980 1991 2001 Spain 1981
1991 2001 Hungary 1980 1990 2001 China? 1982
1990 2000
2nd Release
6
INTEGRATES
IPUMSi
StandardUN/Eurostat Principles Recs...
Census documentation compiled for the Colombia
project
Photos of the IPUMSColombia teamFebruaryMarch
20005 statisticians from DANE 8 academics 3
universities
7
DISSEMINATES
IPUMSi
International webbased access system
EndUser license agreement protects
confidentiality assures proper use User selects
countries censuses cases variables
samplesassists comparative research Open
architecture software
8
Calibration testMexico 1990 2000 FLFP census
microdata vs. employment survey
Employment surveys date from the 80s many
probing questions finely tuned instrument
collected by trained interviewers urbanlack
national coverage Census microdata strength
national coverage back to 1960 weakness
untrained interviewers one questionomits many
working women particularly informal
workers Purpose of paper calibrate census
microdata w/ employment surveys
9
Table 1. Selected microdata samples of Mexico
1960 2000 Year Type Sample Size Density
1960 Census 502702 1.5 1970
Census 480265 1.0 1980 Census No sample
available due to earthquake damage
1990 Census 802774 1.0 1990 ENEU
survey 172233 0.2 2000 Census
10099182 10.0 2000 ENEU
survey 562471 0.6 to be integrated in
IPUMSi
10
The problem table 2 Mexicos global female
labor force participation rate 1264
years microdata 1990 2000 survey ENEU
34.6 43.3 national census 20.6 32.9 ?
14.0 10.4
A better comparisonControl for survey ENEU
sampling frame 16 cities in 1990 survey 16
cities 34.6 41.7 census 16 cities 29.0
40.2 ? 5.6 1.5 includes responses
from LFP questions 1 activity 2
verification.
11
2000 censustwo questions on LFP 1 Last
week did NAME... ...
2000 censustwo questions on LFP 1 Last
week did NAME...? 2 Besides ... did
NAME...?
Question 1 Last week Name Did you work?
27.5 Had work? 0.4 Look for work? 0.3 Are
you a student? housewife? retired?
permanently incapacitated? Did you not work?
Question 1 Last week Name Did you work?
27.5 Had work? 0.4 Looked? 0.3 Q.
12 combined student/wrkd 0.5 housewife/wr
3.7 retired/wrkd 0.0 other/w? 0.4
no reply/w? 0.0
Question 2 Did you help in a family
business? Sell some product? Make some product to
sell? Help on a farm or with livestock? Or in
exchange for pay did you do some other activity?
12
Table 3. Urban Females 1990aged 1264
Structure FLFP Rates
Survey Census Srvy Cnss Total
62248 63929 34.6 29.0 Education Less than
6 years 20.9 21.7 29.3 20.1 Completed
primary 34.7 34.8 27.6 21.1 Completed
middle 20.4 24.3 31.3 37.9 Postmiddle
10 23.9 19.3 53.1 42.2 Marital
Status Married all forms 48.2
50.4 27.7 21.3 Not in union 51.8
49.6 41.4 36.9
13
Table X. Females 2000 Urban same 16 cities
as in ENEU 1990 table not in paper
Structure FLFP Rates
Survey Census Srvy Cnss Total 124051
1073222 41.7 40.2 Education Less than 6
years 14.9 15.9 35.3 31.0 Completed
primary 30.8 28.7 32.1 30.1 Completed
middle 28.5 19.7 47.0 41.0 Postmiddle
10 25.8 36.0 51.1 51.9 Marital
Status Married all forms 51.9
52.9 35.1 34.0 Not in union 48.1
47.1 48.9 47.3
14
Table Y. Logistic Regression Source Females
1990 2000 same 16 cities as in ENEU 1990
table not in paper
Female labor force participationENEU
indicator vs. CensusVariable B
S.E. Wald df Sig R
ExpB 1990 .2212 .0135 267.2291 1
.0000 .0412 .8015 2000 .0860
.0067 163.1781 1 .0000 .0100
1.0898 Model source effect taking into account
age marital status and education.
Interpretation If for both sources weights are
correct and slight structural differences are
taken into account 1990 census underreported
20 of FLFP. 2000 census overreports FLFP by
9.
15
LFP by sex and marital statusMexico 1990 and
2000 national figures
Females
Males
1990
2000
16
Marriage and education strongly affect FLFP
Mexico 1990 and 2000 national figures
Not
Married
1990
2000
17
Figure 3. Female labor force participation rates
by ageUnited States 18801990 and Mexico 1970
1990 2000
18
Conclusions
Mexican census microdata may be more informative
even about FLFP than researchers think Mexican
census microdata on FLFP display remarkable
coherence in time and space Chorus of calamity
on Mexican FLFP may overlook enormous changes in
education weakening power of patriarchy over
married women real advances of women in the
workforce 2000 microdata tell the
story Calibration helps weigh strengths and
weaknesses of sources
19
paper is on conference CDor contactrmccaa@umn
.edu Thank you