2. Motivation: Extreme Lumpiness in
Exporter Data
No. of firms
MedianSS/
MedianNSS
AverageSS/
AverageNSS
SS
Share
SS Man.
Share
Average top 1% 55 1459 164 53 54
Median top 1% 28 1195 108 53 52
Average 1 32458 586 14 17
Median 1 8298 258 11 12
Average 10 7011 283 39 42
Median 10 2171 138 36 41
3. Exporter Concentration Increasing in
Stage of Development
ALB
BELBFA
BGD
BGR
BRA
BWA
CHL
CMR
COL
CRI
DOM
ECU
EGY
ESP
EST
GTM
IRN
JOR
KEN
KHM
LAO
LBN
MAR
MEX
MKDMLI MUS
MWI
NER
NIC
NORNZL
PAK
PER
SEN
SLV
SWE
TZA
UGA
YEM
ZAF
.2.4.6.8
1
ShareTop1%
6 7 8 9 10 11
Ln GDPpc
Share of Top 1% of Exporters - lnGDPpc
4. Motivation Continued
• Important Implications
– Individual firms matter, as in Gabaix (2011), shocks to
individual firms can have aggregate effects, export
growth and diversification
– Individual firms affect type of exports
(effectively, comparative advantage)
• Firms affect technology and RCA can be dependent on a
handful of firms.
– Greater role for oligopoly and FDI—not a draw, but
firms bring productivity from abroad (Neary 2009).
– Fixed versus variable costs and welfare- Entry costs
matter much less that variable costs since infra-
marginal firms are small (di Giovanni and Levchenko
2010)
• Where do big firms come from?
– If big firms drive exports, where do they come from?
5. Export Superstars
• Large exporters (Superstars) define exports
(Freund and Pierola, 2012).
– Superstars drive export growth and diversification.
– Superstars drive sectoral export patterns and
comparative advantage in some industries.
– Superstars are born relatively large and grow fast
• MENA’s export performance has been very
weak. What is the role of MENA’s superstars in
explaining it?
6. Export Superstars
• Large exporters (Superstars) define exports
(Freund and Pierola, 2012).
– Superstars drive export growth and diversification.
– Superstars drive sectoral export patterns and
comparative advantage in some industries.
– Superstars are born relatively large and grow fast
• MENA’s export performance has been very weak.
What is the role of MENA’s superstars in
explaining it?
– Punchline: Mena’s elite export team is in need of a
deeper and more dynamic bench.
8. Data used for this paper: Exporter-level information
on non-oil exports for 32 countries for period 2004-
2010
9. Exports are Defined by a Few Big Firms
All countries MENA (blue) vs. Non-MENA (red)
012345
16 17 18 19 20 21
Ln(size exporter 2008)
012345
lnr
16 17 18 19 20 21
Ln(size exporter)
In MENA, top firm is relatively larger
given the distribution of the other
firms.
Relationship is linear. Distribution
follows Zipf’s law.
Ln (average size 2008) Ln (average size 2008)
Ln (rank)Ln (rank)
10. MENA’s Elite Compared
Number of
Superstars
Ratio
MedianSS/
MedianNSS
Ratio Top
Exporter/MedianNSS
Ratio Median Top
10/MedianNSS
(1) (2) (5) (8)
52 572 9,540 2,055
28 1,195 8,298 2,171
SS = Top 1% of Exporters SS = Top Exporter SS = Top 10 Exporters
All Median
MENA Median
Region
MENA vs. All countries: Top exporter is relatively bigger in MENA.
If taken as a group (top 1% of firms), MENA’s superstars are relatively
smaller.
12. Growth Accounting
• Superstars contribution to overall growth
• Superstars contribution to diversification
– Last 3-consecutive-year period available for each
country
– Intensive margin: all export flows at the country-
HS 6 digit that existed in both Year 1 and Year 3
– Extensive margin: export flows at the country – HS
6 digit that disappear or appear in Year 3 (with
respect to Year 1)
13. Superstars: Export Growth and
Diversification
Int NSS (1) Int SS (2) Ext NSS (3) Ext SS (4) Total NSS (5) Total SS (6) Total Growth (7)
Median MENA 0.22 0.66 0.02 0.10 0.24 0.76 40.96%
Median SSA 0.28 0.50 0.13 0.17 0.36 0.64 46.39%
Median LAC 0.34 0.60 0.04 0.02 0.39 0.61 27.69%
Median ROW 0.62 0.32 0.05 0.04 0.66 0.34 29.76%
Median ALL 0.37 0.48 0.06 0.07 0.44 0.56 34.46%
Median MENA 0.37 0.29 0.10 0.04 0.69 0.31 40.96%
Median SSA 0.75 0.04 0.22 0.00 0.91 0.09 46.39%
Median LAC 0.85 0.03 0.04 (0.00) 0.97 0.03 27.69%
Median ROW 0.86 0.01 0.09 0.01 0.96 0.04 29.76%
Median ALL 0.82 0.05 0.10 0.01 0.92 0.08 34.46%
Median MENA 0.30 0.58 0.03 0.13 0.33 0.67 40.96%
Median SSA 0.51 0.34 0.15 0.16 0.64 0.36 46.39%
Median LAC 0.60 0.37 0.04 (0.00) 0.64 0.36 27.69%
Median ROW 0.76 0.23 0.07 0.02 0.77 0.23 29.76%
Median ALL 0.55 0.31 0.07 0.04 0.65 0.35 34.46%
Top Firm
Top 10
Share in Total Growth
Top 1%
0.44 0.56
Superstars contribute to more than half overall growth across countries
0.69 0.31
Top firm is particularly notable in MENA
0.37 0.48 0.06 0.07
They also explain more than half of growth in both margins of trade.
0.92 0.08
0.10 0.04
0.10 0.01
18. Superstars were already big a decade
ago, or they are new firms
Over 50 percent of 2010’s Moroccan superstars were already so in 2002.
Sense of less dynamism among top firms in Morocco.
19. Strong staying power of exporters in
Morocco
There is little transition of firms to other parts of the firm size distribution
Numbers in diagonal for Morocco are in all cases higher than corresponding
numbers for Costa Rica and Peru
Firm
size
distrib.
in 2002
Firm size distribution in
20. One Firm Can Change
Comparative Advantage
Costa Rica Morocco
Electrical
Machinery Misc. Paper
Electrical
Machinery
Reversal year 2005 2006 2008 2005
Revealed Comparative Advantage
RCA start 0.45 0.44 0.55 0.81
RCA at Reversal 1.27 1.08 1.13 1.01
RCA end 1.24 1.04 1.13 1.09
Revealed Comparative Advantage, excluding the top firm
RCA start 0.45 0.32 0.32 0.39
RCA at Reversal 0.63 0.67 0.49 0.66
RCA end 0.49 0.75 0.49 0.88
Top Firm
Export share at start 0.01 0.29 0.43 0.55
Export share at reversal 0.56 0.41 0.57 0.39
Export share end 0.65 0.30 0.57 0.22
21. Conclusions
• Importance of top similar in MENA to the rest of the world.
– Single largest firm in MENA tends to be relatively large as compared
with countries in other regions. The other top firms are not as
outstanding.
– MENA superstars contribute more to growth and diversification,
especially top firm.
– MENA superstars contribute less to creating new comparative
advantage sectors.
– Some evidence of less churning among MENA’s superstars.
• Consistent with evidence from private sector studies
• Relatively old managers
• Relatively old firms
=> Weak lifecycle dynamics in MENA, not enough exporters able to
grow into large global exporters and break new industries.
22. Policy Implications
• Variable rewards/costs matter more than fixed
costs
– Exchange rate is critical—overvaluation has huge
costs.
– Trade costs, tariffs and customs are important
– Business climate must not impede firm growth
– Trade agreements
• SME financing is not a viable tool for export
growth