IPUMS-International Disseminatin

Document Sample
IPUMS-International Disseminatin Powered By Docstoc
					           Engendering economic activity in
                population censuses
    Robert McCaa, Antonio Lopez, Phuong Nguyen
       University of Minnesota Population Center
    * * * representing the IPUMS-International * * *
      * * * census microdata access project * * *
 * currently available, free: 80 censuses, 26 countries *
* * 75 countries have entrusted microdata to project * *
         Engendering economic activity—
   a global perspective acquired from integrating
                 census microdata
• IPUMS: most complete archive of census documentation
   –Census forms: 765 on conference DVD—thanks to UNSD, etc.
   –Training manuals: available in 2009
   –To integrate microdata, study documentation carefully
• “best practices” for capturing female economic activity
   –Learned from source documentation
   –Necessary for integrating microdata
   –And writing integrated metadata
• (Regional reports, on Latin America and Africa given in May and
Dec., provoked much discussion and debate)
      UN-Statistics Division, Principles and
    Recommendations, 1958-2007: conceptual
foundation for unbiased measurement of econ. act.
• ~International standards are clear:
  homemakers may be economically active
  – P&R 1958 (M27), 1970 (M44), 1998 (Rev 1, M67), 2007 (Rev 2)
  – UNSD Soc. & Econ. Char. Handbook (2007), ILO BLS 2007-1
• Biased questionaire:
   Requests “main activity”; if housewife, end module.
• Unbiased:
   Requests any economic activity; if housewife,
    continue with a short list of probes:
  –Work for pay or profit, family farm/business, crafts for sale,
  –For list: see UNSD Handbook (Nov. 2007), p. 45.
     UN-Statistics Division, Principles and
Recommendations, 1958-2007, laid the foundation;
   yet our analysis of 765 census forms shows
  –before the 2000 round, biased wording meant that
  homemakers/housewives who worked part-time were excluded
  from the economically active population
  –Finally, in the 2000 round, in almost all regions of the world,
  bias was greatly eliminated—thanks to a concerted effort by
  UNSD, UNFPA, etc.
      Examples: Mexico, South Africa, Macedonia, Yemen
  –For 2010 round, almost all (?) countries may pose the economic
  activity question properly (e.g., Colombia 2005, Cambodia 2008),
  but, unless we act quickly, old ways will continue in much of
  Africa (e.g., Nigeria 2006, Burkina Faso 2006)
  and sub-regions of Asia (Indonesia, Thailand, Vietnam)
 Gambia 2003—
  captures both
   primary and
secondary activity
• Homemakers:
must also answer
16b (“did you work
at all for pay or
profit”, work for
family farm,
business, receive
pay in kind)
•If yes, continue
with economic
 Mexico 2000 census--two questions on Economic Activity:
    1: “Last week, did (NAME)...?”
   2: “Besides (...), did (NAME)...?”
Question 1: female % Question 2:
Last week (Name):    Did you help in a
Did you work? 27.5% family business?
Had work?        0.4 Sell some product?
Look for work? 0.3 Make some product
Q. 1&2: a combined
Are you                to sell?
 student/wrkd 0.5 Help on a farm or
 housewife/wr 3.7
   housewife?          with livestock?
 retired/wrkd 0.0 Or in exchange for
   permanently 0.4     pay did you do
 no reply/wrkd? 0.0
  incapacitated?       some other
Did you not work?
      Yemen 2004
one question, 2 categories:
    1: “Working”
2: “Working housewife”
      South Africa 2001—captures all with short list:
“any work for pay (in cash or kind), profit or family gain
 for one hour or more?” (avoids “homemaker deadend”)
    The problem:
 Nigeria 2006—skips
  secondary activity
•17: if Homemaker,
skip: “end interview”.
  –17b: list secondary
  activities (recall Mexico
  2000 example)
  –If response is “no” on
  17a and 17b, then end
  interview; otherwise
  record answers for
  17b, 18 and 19
      1.00 to 2.00
      2.00 to 3.00              None
Gender bias in the wording of
      3.00 to 4.00              Some
      4.00 to 5.00              Substantial
Census Questions on Economic
      5.00 to 6.00
      6.00 to 6.00
                                All, if homemaker
                                No census/ques/form
     Activity (2000phcr)
      No Data
      All Others
        Africa, 1995-2007—3 of 4 miss most or all
      economic activity of homemakers: 35 censuses
6 best practices: Gambia 2003, Mauritius 2000, Mozambique 1997,
          Namibia 2001, South Africa 2001, Sudan 2008

   3 capture some (all??) secondary economic activity:
           Cote d’Ivoire 1998, Ghana 2000, Mali 1998
        9 miss most homemaker economic activity:
 Benin, Cameroon, Egypt, Ethiopia, Guinea, Morocco, Seychelles,
               Sierra Leone, Swaziland, Tunisia

   17 skip all female economic activity, if homemaker:
     Algeria, Burkina Faso, Botswana, Cape Verde, Gabon,
    Guinea (Conakry), Kenya, Lesotho, Malawi, Mauritania,
               Niger, Nigeria, Senegal, Tanzania,
                 Uganda, Zambia, Zimbabwe
     Asia-Oceania, 1995-2007—half omit some or all
      economic activity of homemakers: 43 censuses
21 best practices: Australia, Azerbaijan, Cambodia, Fiji, Georgia,
Hong Kong, India, Iran, Israel, Japan, Jordan, Korea-RO, Malaysia,
    Maldives, New Zealand, Palestine, PNG, Sri Lanka, Syria,
                      Turkmenistan, Yemen
    8 capture some (all??) secondary economic activity:
   China, Kazakhstan, Kyrgyzstan, Nepal, Pakistan, Singapore,
                       Taiwan, Turkey
1 misses much female economic activity for lack of probes:

    14 skip all female economic activity, if homemaker:
 Bahrain, Bhutan, Indonesia, Iraq, Kuwait, Laos, Mongolia, Oman,
    Philippines, Saudi Arabia, Thailand, Timor Leste, Vietnam
        Europe, 1995-2007—half omit some or all
      economic activity of homemakers: 31 censuses

15 best practices: Austria, Belgium, Bulgaria, Croatia, Denmark,
France, Italy, Macedonia, Netherlands, Norway, Poland, Portugal,
                  Russia, Switzerland, Ukraine

   5 capture some (all??) secondary economic activity:
        Estonia, Greece, Hungary, Spain, United Kingdom

4 miss much female economic activity for lack of probes:
               Albania, Belarus, Ireland, Moldova

    7 skip all female economic activity, if homemaker:
    Bosnia & Herzegovina, Czech Republic, Latvia, Lithuania,
                  Romania, Serbia, Slovenia
   Americas, 1995-2007—only 6 less than best; little
gender bias in economic activity question(s): 29 censuses
 23 best practices: Argentina, Bahamas, Belize, Bolivia, Brazil,
 Canada, Colombia, Dominican Republic, Ecuador, El Salvador,
Guatemala, Guyana, Haiti, Honduras, Jamaica, Mexico, Nicaragua,
  Paraguay, Puerto Rico, Suriname, USA, Uruguay, Venezuela

   0 capture some (all??) secondary economic activity:

2 miss much female economic activity for lack of probes:
                         Panama, Peru

    4 skip all female economic activity, if homemaker:
          Chile, Costa Rica, Cuba, Trinidad and Tobago
   UNSD, Handbook on Social and Economic
Charactertistics (Nov. 2007, p. 45): list of economic
           and non-economic activities

      Continued on next slide
Handbook, 2007, p. 45:
 economic activities

     Continued from previous slide
    A critique of examples from Handbook
                 (pp. 310-340).
–3 Worst practice (if main activity = homemaker, skip):
Botswana 2001, Malawi 1998, Zambia 2000.
–13 Best practice (record activity, even for
homemakers): Namibia 2001, South Africa 2001, Sri
Lanka 2001 (weeks worked per year for 6 types of
activity, including homemaking), Argentina 2001,
Bahamas 2000, Belize 2000, Jamaica 2001, Mexico 2000,
Australia 2006, Japan 2000, Canada 2006, France 1999,
and Italy 2001
–There is no guidance as to the character of these
examples—instead only layout (landscape/portrait) and
# of questions asked.
Suggested “best practice” on economic activity proposed
 for Africa Addendum to UNSD P&R 2010, Option 1:

                     Question X (mark one):
                     Last 7 days, did (Name) do any
                         work for PAY (in cash or in
                         kind), PROFIT or FAMILY
                         Gain, for one hour or more?:
                     1. Yes, formal (non-farm)
                     2. Yes, informal (non-farm)
                     3. Yes, farming/ranching
                     4. Yes, has work, but was
                         temporarily absent
                     5. No, did not work.
                     If No, skip to next module.
  Suggested “best practice” on economic activity
 proposed for Africa Addendum to UNSD P&R 2010.
Option 2: Branch only after 2nd (probing) question
         X: “Last [period], did (NAME)...?”
       X+1: “Besides (...), did (NAME)...?”
Question X (check one):    Question X+1 (check one),
Last week, did (Name):      Besides (…), did (Name):
1. Work? >>Q Z.            1. Help in a family business?
2. Have work? >>Q Z.       2. Sell some product?
3. Look for work? >>Q Z.   3. Make some product to sell?
4. Go to school?           4. Help on a farm or with
5. Perform household         livestock?
  tasks?                   5. Do some other activity in
6. Do other things?          exchange for pay?
                           6. None of above >>skip to
                             next module
 Suggested “best practice” on economic activity
proposed for Africa Addendum to UNSD P&R 2010:
             Option 3. Add no question—
   only a few additional words and do not skip!
Question X (check one):
1. Worked for pay/profit/family gain.
2. Had work, but didn’t work.
3. Looked for work?
4. Went to school; didn’t work at all.
5. Performed household tasks; didn’t work at all.
6. None of the above.
Partners: IPUMS-International, December 10, 2007
dark green = disseminating (26 countries, 80 censuses, 200mpr)
  green = harmonizing (41 countries, 112 censuses, 230mpr)
                 lightest green = negotiating

                           Mollweide projection
     IPUMS Dynamic Metadata System:

 2 “clicks” to view images of original source
        documentation—census forms
and training manuals—in the official language
            (or English translation)
For images of original source documents,
     2 clicks from IPUMS home page:
     click 1: “Census Questionnaires”
For images of original source documents,
     2 clicks from IPUMS home page:
     click 1: “Census Questionnaires”
 click 2: country, census year, document
    IPUMS Dynamic Metadata System:

 5 “clicks” to compare any census question,
in English, for any combination of years and
          countries in the database
Click 1, access:
Click 2: “variables”
Click 3: Select countries and census years;
       then, submit sample selections
Click 4: chose a variable (e.g., employment status
            —i.e., economic activity)
        Before making the fifth click,
scroll down to see comparability discussion
Click 5 “enumeration text” to compare text of enumeration
       form and instructions for Employment Status
Click 5: “enumeration text”
For variable availability, click variable name
  (X = variable available for this census)
For sample counts, click case count view
 (careful: these are unweighted counts)
For simple variables, click “general codes”
           Thank you

 additional information at:


Shared By: