Data Hunting in Brazil

1
Data Hunting in Brazil

• Adalberto Cardoso (coord - IUPERJ)
• Alberto Najar (FIOCRUZ)
• Miguel Murat Vasconcellos (FIOCRUZ)
• Jacques Levin (FIOCRUZ)
• Silvia Rangel (FIOCRUZ)
• Carlos Antônio Costa Ribeiro (IUPERJ)
• Glaucio Ary Dillon Soares (IUPERJ)
• José Ricardo Ramalho (UFRJ)
• Luiz Cesar Queiros Ribeiro (IPPUR-UFRJ)
• Celi Scalon (UFRJ)

Sponsored by the ESRC, conducted at IUPERJ -
Brazil
International Data Forum
2
(No Transcript)
3
(No Transcript)
4
(No Transcript)
5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
Placed in the New World

• Brazil is as big as a Continent (more than 8
  million square km and 8 thousand km of beaches)
• 190 million people live there, 85 of which in
  urban areas 36 of which in metropolitan areas
• São Paulo Metr. Area alone hosts 10 of the
  countrys population

13
Brazil is a federation of states
14
This makes it very hard to centralize data
production

• First, because many public policies are designed
  at the local (state or municipality) level. So,
  many states and cities and metropolises have
  their statistics bureaus
• Many administrative registers (specially on
  violence ever, and health before 1988) are
  gathered at the local level and centralized by a
  federal agency
• Concern for data validity and reliability is
  increasing but there is still much to do

15
At the Federal Level, IBGE is the main data
producer and disseminator

• It is also responsible for the standardization of
  socio-economic classifications (occupations,
  economic sectors, product denominations and many
  others)

• Celebrating 70 years
• It gathers data on population, economy (including
  national and regional accounts), labour market
  and many more

16
The main IBGE surveys are

• The Monthly Employment Household Survey (PME) in
  6 metropolitan regions
• Household budget surveys

• National Decennial Census Survey
• Annual National Household Sample Survey (PNAD),
  80 thousand households and 400 thousand
  individuals surveyed every year

17

• And Annual and Monthly Economic Surveys on
  Manufacturing, Services, Commerce, Construction
  and Agriculture
• IBGE also hosts data produced by the ministries
  of health and justice (violence data)

18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
To access this information all you need is

• A good internet connection

If you want more sophisticated analysis
Then you can buy the CDs or access BME at a very
low price
23
IBGE has a stable staff of public servants
geographers, economists, statisticians,
sociologists, historians, biologists and many
others. But it does not produce in depth
socio-economic analysis
24
It relies on the academia for this matter.In
fact, the legitimacy of its methodologies and
procedures is built in close cooperation with
academic researchers.For instance, deep changes
in PNAD and PME are being discussed at this
precise moment with experts from more than 30
research institutions
25
The main problems

• All the material is in Portuguese
• Constant changes in methodology makes it
  difficult (but not impossible) to compare surveys
  over time
• Standardization of old surveys is poor, one needs
  to get through documentation very carefully, and
  cooperation with local experts is inescapable

26
There are some important academic institutions
that host hundreds (sometimes thousands) of
surveys on all kinds of issues, from violence to
political values to demography and
consumption.The most important is CIS
(Consortium of Social Information), sponsored by
the ANPOCS(www.cis.org.br)
27

• The problem here is of a different sort
• We still have to create a culture of quantitative
  social analysis. Most research programs in
  leading universities are qualitative
• So, it is difficult to convince the peers (who
  are peer reviewers for funding agencies) that
  money should be directed to expensive
  quantitative research or archive, the results of
  which they do not believe in