Universidad Nacional de General Sarmiento

1
Universidad Nacional de General Sarmiento

• Lessons from the Buenos Aires Time Use Survey
• A Methodological Assessment
• Valeria Esquivel

8th International Summer Conference on Knowledge
Networking and Engendering Macroeconomics, Gender
and International Economics Istanbul, Turkey,
20th July 2007
2
The Latin American Experience in TUS data
collection

• A cumulative process of learning by doing since
  first experiences took place (Cuba in 2001,
  Mexico in 1996, Nicaragua in 1998). A new breed
  of TUS comprising the 2006 Ecuadorian TUS, the
  2003 Uruguayan TUS and the 2005 Buenos Aires TUS.
• Different approaches to TUS according to
  objectives and budgetary restrictions.

3
The Latin American Experience in TUS data
collection

• Strong advocacy for modular approaches, though
• There is an excessive emphasis in advocating for
  tasks list surveys, even though their drawbacks
  are clearly recognized
• based almost exclusively on costs
  considerations.

4
The BA TUS Institutional framework

• Local legislature mandate to collect time use
  data on a regular bases to measure womens
  contribution to the economy.
• Cooperation agreement between Universidad
  Nacional de General Sarmiento and the
  Directorate-General of Statistics of the Buenos
  Aires City Government.

5
The BA TUS as a module of AHS

• The Buenos Aires Annual Household Survey is
  collected annually by the DGEyC. Multi-purpose
  household survey, representative of Buenos Aires
  population (8 of country).
• Data on social and demographic variables
  households and families dwellings health
  education labor market status incomes social
  policies.
• The BA TUS was a module applied to 1/3 of total
  surveyed households.

6
Main methodological features

• a. Survey type module attached to an ongoing
  household survey
• b. Data Collection Instrument Activity Diary
• The activity dairy started 4AM yesterday and
  lasted till 4AM of the interview day. The diary
  was a closed grid consisting on 48 30-minutes
  long time slots that allowed for up to three
  consecutive and/or simultaneous activities. Every
  activity had a location mark.

7
Main methodological features

• c. Data Collection Method 24 hour recall
  interview
• This diary was filled in by interviewers (not
  respondents) using a series of questions that
  prompted respondents recall.
• Respondents reported in their own words their
  previous days activities, starting on their
  waking up time and ending when the respondent
  went to sleep.
• A special set of questions was designed to ease
  the reporting of paid working hours and at the
  same time avoid missing simultaneous activities
  performed while doing paid work.

8
Main methodological features

• d. Unit of analysis the individual
• Only one randomly selected household member,
  either women or men between 15 and 74 years of
  age answered to the diary.
• The BA TUS respondent might coincide (or not)
  with the core survey respondent. Like many other
  multi-purpose household surveys, one adult
  household member could answer the whole of the
  Annual Household Survey, but this member might or
  might not be chosen to answer to the time use
  module.

9
Main methodological features

• e. Contextual variables
• Location codes.
• f. Probing questions
• Probing questions were added to check whether
  some of the activities mentioned were done for
  pay (thus helping post coding) and to capture
  child care time which goes frequently
  underreported. A question asking on whether the
  day reported was typical was also added to act as
  control.

10
Probing questions

11
Main methodological features

• g. Treatment of simultaneous activities
• Up to three non-hierarchical simultaneous
  activities in each time slot

12
Main methodological features

• h. Activity Classification System
• Aggregation issues in relation to the ACS
  (2-digit).
• Only one SNA-work category was used (100, and
  200 and 300 activities were not included), since
  very few (if any) primary production for self
  consumption takes place in the Buenos Aires City
  urban context.
• Caring for children activities and special
  cases.

13
Activity Coding

• MANY ACTIONS CAN BE AN UNIQUE ACTIVITY,
  REGISTERED ON A LINE

14
Activity Coding

• MANY ACTIVITIES COULD BE SUMMARIZED IN A SINGLE
  ACTION

15
Some quality indicators

• 80 of interviews took 20 minutes or less.
• 68 of overlapping between AHS respondent and TUS
  respondent.
• Household member respondent highly correlated
  with size and household type.
• 10 non-response rate.
• 21 average episodes.
• 28.3 hours of avarage day when simultaneous
  activities are fully accounted for.