Measuring U.S.-Mexico Remittances

1
Measuring U.S.-Mexico Remittances

• Pia M. Orrenius
• Joint work with Jesus Cañas Roberto Coronado

Disclaimer The opinions are those of the
presenter and do not reflect the views of the
Federal Reserve Bank of Dallas or the Federal
Reserve System.
2
Puzzling data

• Official remittances
• High, fast-growing
• 20 billion in 2005
• About 2,800 per Mexican worker in U.S.
• About 13 percent of pre-tax earnings
• 25 percent average annual growth since 2000
• Tripled since 2000 (6.6 billion)
• Series more volatile

3
U.S.-Mexico remittances surge2005 dollars
Source Banco de Mexico
4
U.S.-Mexico remittances surge2005 dollars
Source Banco de Mexico
5
while in-migration declines Estimates of annual
migration to US by legal status
Source Jeffrey Passel and Roberto Suro, Pew
Hispanic Center (2005)
6
Puzzling Data Why Should We Care?

• Important the numbers be right
• Home country
• Household income, consumption
• Poverty estimates
• Social programs
• Macro effects
• Host country
• Capital outflows, development assistance, FDI
• Banking access, regulation
• Cost-benefit of migration policy reform

7
Contradictory data

• BEA remittances 10.2 billion in 2005
• About 1,460 per Mexican worker
• Model-based estimate
• Banco de México uses bank reporting, surveys
• Recipient-based micro data (Esquivel, 2005)
• About 2,600 per household (ENIGH 2002)
• 1.4 million recipient households
• Adds up to 3.6 billion in 2002 (37 total)
• Sender-based micro data
• 79 percent Mexican workers remit 350/month (MMP
  transfer data)
• Adds up to 23.2 billion but
• MMP is not representative of Mexican immigrants
  in U.S.

8
BEA, Banxico differencessurge after 2002
(U.S.-Mexico remittances)
9
BEA, Banxico differencesurges after 2002
(U.S.-Mexico remittances)
10
BEA, Banxico differencesurges after 2002
(U.S.-Mexico remittances)
11
Model of remittances

• ARMA model of quarterly remittances on its
  drivers for 1980q1 to 2006q2
• Explanatory variables include
• United States real gross domestic product
• Mexico real gross domestic product
• Peso/dollar exchange rate
• United States consumer price index
• United States unemployed
• Maquiladora employment
• Autoregressive parameter
• Moving average parameter

12
Our model cannot predict the post-2000 surge
(U.S.-Mexico remittances)
Source Banco de Mexico
13
Our model cannot predict the post-2000 surge
(U.S.-Mexico remittances)
Source Banco de Mexico Authors calculations
14
Our model cannot predict the post-2000 surge
(U.S.-Mexico remittances)
15
Remittances to Developing Countries
Source IMF BOP
16
Remittances to Developing Countries
Source IMF BOP
17
Mexico, India outliers?
Remittances to Developing Countries
Source IMF BOP
18
Mexico, India outliersor not?
Remittances to Developing Countries
Source IMF BOP
19
Some reasons to expecthigher remittances

• Better measurement
• Banco de México changed accounting methodology in
  2002
• Less cash and carry
• Less circular migration, longer stays
• Heightened scrutiny of return migrants
• Transfer costs have fallen
• Peso has remained strong

20
Conclusion

• Work in progress
• More questions than answers
• Difficult to reconcile different data sources
• Implications broad for home, host country
• Stay tuned!

21
(No Transcript)