Transcript slides

Comments on: “Determinants and effects of
export diversification in Cape Verde and
Mozambique: A contribution towards
assessing successes in West and Southern
African economic communities,”
by Jorge Braga de Macedo, Luis Brites Pereira
and Manuel Caldeira Cabral
Jeff Frankel, Harvard Univ. & NBER
For Economic Research on African Development
Successes Pre-Conference, Feb. 28, 2009
Historical background
• Most authors at this conference weren’t born
when the countries they are studying became
independent.
• But Jorge was actually a colonial occupier !
• When I first met him in Lisbon in 1976, he was a
junior army officer, just-returned from Angola.
• But then, junior officers were the ones who
turned against the colonial wars and overthrew
the Portuguese dictatorship. So that’s okay.
2
Overall…
• Compliments:
– I learned a lot from reading the paper.
– It blends econometrics with country cases.
– It raises some interesting questions
• Grumble, grumble:
– I only got the current version Friday.
– It is 64 pages long, single-spaced.
• Its title is almost as long as other people’s papers here!
– It is really two different papers, with little connection:
• 1st half is econometrics
on Export Diversification & Export Sophistication
• 2nd half is indicators for Cape Verde & Mozambique
3
Why the choice of Cape Verde & Mozambique?
• Obviously they are Portuguese speaking,
but why these two out of the five?
• Cape Verde? It counts as a success.
• Mozambique? It recovered rapidly after its civil war.
• I would propose perhaps re-casting this half
of the paper as two comparisons:
– Cape Verde versus Sao Tome & Principe (2 small island countries)
– Mozambique vs. Angola (2 remote mainland countries each of
whom fought a long civil war).
– The question then would be why Cape Verde has done
better than ST & P and Mozambique over Angola.
– If the question were phrased this way, the obvious starting point
4
would be Natural Resource Curse: both ST & P and Angola have oil.
Indicators of governance (rule of law…)
and outcomes (income per capita, human development…)
• Part 3 of the paper shows lots of indicators
from the usual sources (World Bank, etc.)
• I will show some summary measures from the
annual Index of African Governance by Rachel
Gisselquist & Robert Rotberg, Harvard.
5
Index of African Governance
(for 2006, published 2008)
1.
2.
3.
4.
5.
…
Mauritius
Seychelles
Cape Verde
Botswana
South Africa
24. Niger
…
44. Angola
45. Sudan
46. Chad
47. Congo (DR)
48. Somalia
Three island countries have the best institutions,
as measured by Index of African Governance
60
70
Mauritius,
Seychelles
SC
&
Cape Verde
are rivaled only
by S.Africa
ZA
GA
& Botswana
50
NA
CG
GQ
KE
SZ
GH
MU
BW
CV
CM
40
SD
CI
30
ER
AO GM
TZ
LS
SN
UG ZM
TG
ST
MW
ETNGGM
MG
Cape Verde
ZW
RW
CC
better than
BJ
MR DJ BI LR MZ
Sao Tome
CF
& Principle
Angola falls far short of Mozambique
SO
CD
TD
SL
ML
BF
NE
20
GW
0
20
40
60
Average of rule of law and human rights
80
7
“Globalization”
• “In less than one year, the risk of excluding African
economies from globalization has turned into the hope
of sparing [them] from the global financial crisis (p.1).”
-- …Very well said !
• “…globalization in the form of export diversification
and sophistication interacts with development and
governance indicators… (p.44) ”
-- …Not so well said.
• Globalization ≠ Export Diversification & Export Sophistication.
– A country that produces only minerals for export is globalized !
• Interacting globalization measures with governance indicators
in a growth equation is a good idea. Why not try it?
8
Regional integration
• The authors present the interesting proposition that
regional integration helps globalization
– Because economies-of-scale helps trade (Helpman-Krugman)
– & regional associations reinforce political will to reform
• Apparently for CFA countries
– But not for ECOWAS & SADC,
• to which Cape Verde & Mozambique, respectively, belong.
• An appealing notion, but they could pursue it further.
– Needs to be demonstrated; versus, e.g., trade-diversion.
9
Export diversification & income/cap
The authors’ Table 2.4.
SUB-SAHARIAN AFRICA - NUMBER EQUIVALENT AND GDP PER CAPITA
8000
Income
per
capita
Oil Exporting
Countries
7000
6000
5000
4000
3000
2000
1000
0
0
5
10
15
20
25
I’m not sure what is supposed to be the theory,
or even the direction of causality !
30
Export
diversification
10
Export diversification & income/cap
The authors’ Table 2.4.
• I’d expect those clustered at lower left to be small countries.
– Theory predicts: small countries have both lower income
& less diversification within exports (despite higher X/GDP).
• Two reasons
– a narrower range of endowments
– Increasing returns to scale.
• Indeed the authors find size a significantly determines ExDiv (Fig. 2.6)
– even more than does income/cap;
– especially within Africa (p. 13)
• Thus I would think that controlling for size would cut
the positive correlation between Ex. Div. & income.
– But they find controlling for size raises the correlation.
(p. 12)
11
Other controls in the Ex.Div. equation include:
• Barriers to trade
– Landlockedness
– Distance
• which turns out to mean distance from Europe.
• As they note, Niger is closer than Namibia or Mauritius
• One can do better
– (1) There exists measures of distance to nearest port, and then by sea.
– (2) Compute Remoteness: GDP-weighted average distance from all trading partners.
– Tariffs
(p. 17)
• All the barriers to trade reduce Export Diversification
(not just X/GDP), as one would probably expect.
12
Measures of governance
raise Ex. Div. & Ex. Soph.
(pp. 17-19)
• I would’ve guessed the effect comes via income
– But they find it holds even controlling for income.
• I would have guessed it is an artifact of countries
highly specialized in minerals (NRC; e.g., Engerman & Sokoloff) .
– But they find it holds even controlling for oil.
13
Ex Div apparently lowers infant mortality.
“Half as important as income”
How? What channel?
Export Diversification and Infant Mortality
300
250
(Label
the axes,
please)
200
150
100
50
0
0
20
40
60
80
100
120
140
160
180
14
Ex Soph apparently lowers infant mortality.
How? What channel?
Export sophistication and Infant Mortality
250
200
150
100
50
0
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
-50
The same question for the relationship with income:
what is the causal direction of the relationship with ExDiv.?
15
Meanwhile,
back in Cape Verde & Mozambique…
• Africa’s Ex. Div. & Ex. Soph. << all countries (stat. sig. P.21)
• Mozambique’s Ex. Div. & Ex. Soph. ≈ Africa’s (insig. diff.)
– But fnote 10 reveals an aluminum project dominates Mozambique’s Ex. Div. ,
apparently driving a bigger wedge between the two halves of the paper.
• Cape Verde’s Ex.Div. & Ex.Soph. >> rest of Africa’s
(stat. sig.)
16
Ultimately, the keys to Cape Verde’s success:
• Highly diversified exports (p. 28).
• Good governance (p. 38) .
• Or is it divine providence?
– The authors refer to
“…its success in having god governance.” (p.21)
17
Addendum: 3 island countries have the best institutions,
as measured by Index of African Governance,
25000
but some oil producers show up with relatively high income
15000
20000
GQ
SC
GA
10000
BW
MU
5000
ZA
SD
0
CD
0
CI
NA
AO
SZ
TD
ER GM
CG
CM
MR
NG DJ
SNGH
KEML
ZMMG
BJ ST LS
CC
GMCF
BF
UG
TG
RW
MZTZ
MW
ET
SL
NE
ZW
GW
BI
LR
20
40
60
Average of rule of law and human rights
CV
80
20