International migration in Venezuela and the Growth Collapse

1
International migration in Venezuela and the
Growth Collapse

• Dan Levy
• Visiting Assistant Professor, Kennedy School of
  Government, Harvard University
  
• Dean Yang
• Assistant Professor, Ford School, University of
  Michigan
  
• Presentation at the Kennedy School of Government,
  April 2006

2
Two key objectives

• Characterize history of international migration
  in Venezuela
  
• Assess link between migration and growth collapse
  in Venezuela
  

3
Data

• Encuesta de Hogares (1975-2003) Semi-annual
  household survey
  
• Allows us to get a history of number of foreign
  born people and net flows
  
• Key variables
• Place of birth
• Year of birth
• Entrepreneur status
• Limitations
• Year of arrival to Venezuela only available for
  1994-2003
  
• Can only infer net flows
• Entrepreneur status variable is relatively crude
  measure
  

4
I - History of International Migration in
Venezuela
5
Key questions

• How has the stock of immigrants evolved over
  time?
  
• How has the composition of immigrants evolved
  over time?
  
• How do immigrants differ from natives in their
• Entrepreneurship
• Human capital

6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
Main take-away points

• Fraction of foreign born about 5
• Went up in 1975-1980 and down afterwards
• Big shift in composition of foreign born from
  mostly Europeans to mostly Colombians
  
• Foreign born much more likely than Venezuelans to
  be entrepreneurs
  
• But fraction of foreign born entrepreneurs
  decreased over time
  
• Foreign born do not differ markedly from locals
  in terms of educational attainment
  

14
II Assessing link between international
migration and the growth collapse in Venezuela
15
European presence and relative economic conditions
Notes -- Per capita real GDP in S. European
sources is weighted average across Spain, Italy
and Portugal. Weights based on relative shares of
foreign-born across those three countries from
1971 census (0.46, 0.28, and 0.26, respectively).
"Entrepreneurs" are those reporting selves as
"Patronos" in labor force survey.
16
Migration and the collapse key questions

• Generally
• How have the large changes in migrant inflows and
  outflows affected the Venezuelan economy in the
  last 30 years?
  
• In particular
• How have changes in the fraction foreign born
  affected Venezuelan natives labor market
  outcomes?
  
• Outcomes wages, employment, labor force
  participation, total output or production
  
• Focus on natives in the same group as
  foreign-born, such as 1) industries, 2) states
  
• Examine differential effects of Colombian vs.
  European migrants
  

17
Challenges to identification

• Fraction foreign born (in industries or states)
  is endogenous with respect to outcomes of
  interest
  
• Reverse causation growing industries may attract
  more foreign-born
  
• Omitted variables third factors (e.g., industry
  subsidies, sector-specific trade liberalization)
  may influence both natives labor market outcomes
  and fraction foreign-born
• Need a source of variation in fraction
  foreign-born that is exogenous with respect to
  natives labor market outcomes

18
Identification strategy

• Use shocks in migrants origin countries as
  instruments for changes in fraction foreign-born
  
• Economic conditions in origin countries should
  affect new inflows, as well as rate of return
  migration
  
• To get variation at industry or state level,
  interact shocks with initial (1975) fraction
  foreign born in industry or state

19
First stage regression

• Unit of observation is group-country-year (group
  is industry or state)
  
• For fraction foreign-born Fjkt in group j from
  country k in year t
  
• Fjkt g0Zkt g1Zkt-1
• a0(ZktFjk1975) a1(Zkt-1Fjk1975)
• year fixed effects
• country k fixed effects group j fixed
  effects
  
• ejt
• Zjt, Zjt-1, are current and lagged economic
  conditions in origin country k
  
• Interaction with initial fraction in group j
  Fjk1975 allows effect of shock to be larger for
  groups with higher initial presence
  
• Construct predicted fraction foreign born,
  Fhatjkt, for each origin country k
  
• Main concern first stage may be too weak

20
Instrumental variables regression

• For labor market outcome Yjt in group (industry
  or state) j and year t
  
• Yjt b0Fraction_Colombianjt
  b1Fraction_Europeanjt
  
• year fixed effects
• group j fixed effects
• ejt
• Fraction_Colombianjt , Fraction_Europeanjt, are
  predicted values from first stage
  
• Coefficients b0, b1, are effect of exogenous
  changes in fraction foreign born on labor market
  outcomes of natives in same group

21
Comments welcome
22
(No Transcript)