Professional Documents
Culture Documents
Mayer Otaviano - The Happy Few The Internationalisation
Mayer Otaviano - The Happy Few The Internationalisation
1007/s10272-008-0247-x
COMPETITIVENESS
Table 1 Figure 1
Share of Exports of Top Exporters in 2003, The Superstar Exporters Phenomenon
Total Manufacturing (France, restricted sample)
Country of origin top 1% top 5% top 10% 100
Figure 2
Fact 1 – Aggregate exports are driven by a small
The Superstar Exporters Phenomenon, Logarith-
number of top exporters. The top 1%, 5% and 10%
mic Transformation (France, Exhaustive Sample)
of exporters account for no less than 40%, 70%
100 and 80% of aggregate exports respectively.
80
percentage of French exports
60 Export Intensity
40 The fact that only a handful of firms drive aggregate
exports suggests that export status is a mixed bag
20 containing different types of firms.
Table 2 shows that the share of sampled firms that
10 export is roughly 65%, 60%, 45%, 75% and 40% for
France, Germany, Hungary, Italy and Norway respec-
tively. The higher percentages for France, Germany
5
.001 .01 .1 1 2 3 45 10 50 100 and Italy reflect the biases of these samples towards
percentile of exporters relatively large firms. For each country the table re-
1998 2003 ports the percentages of firms exporting more than
the given shares of their turnover, and the percentages
to the case in which all firms export the same value. of total exports accounted for by these groupings of
Hence, the further away the curve is from the line, the firms.
more concentrated aggregate exports are in the hands The results for France, Italy and Norway are similar.
of a few firms. Using the restricted sample, we can plot They show that, even though only a small subset of
a similar distribution for employment (in black) as an firms exports a major share of their turnover, they still
interesting benchmark. Figure 1 shows that the con- account for a large fraction of total exports. In France,
centration is high in terms of employment (the black Germany and the United Kingdom, around 10% of all
curve is far from the uniform distribution), but is much firms export more than 50% of their turnover but they
higher in terms of exports. account for 50% to 75% of total exports. The distribu-
In addition Figure 2 zooms onto the contributions of tion can, however, vary substantially across countries.
“superstar” exporters by showing what happens with- In this respect, an interesting comparison between
in the club of the top 1% of exporters.1 The picture is France and Germany exemplifies the potential of firm-
again striking: the top 0.001%, 0.01% and 0.1% of ex- level data analysis. Germany has a larger proportion
porters still account for not much less than 10%, 20% of firms exporting more than 50% of their turnover,
and 40% of aggregate exports respectively. and they represent a much larger share of total ex-
For Europe in general, we can summarise the find- ports than in France. From Table 2 we can see that
ings as: the greatest contribution (68%) to total exports in Ger-
many comes from firms exporting from 50% to 90% of
1
As we focus here on a smaller number of firms, we need to use the
exhaustive sample to obtain a representative distribution. The logarith-
their turnover. In France on the contrary, the greatest
mic transformation is used to enhance the readability of the picture. contribution (46%) comes from firms exporting from
Table 2
Distribution of our Sample of Exporters by % of Turnover, 2003
% firms exporting more than % of total exports by firms exporting more than
total mfg
exports % export- 5% of 10% of 50% of 90% of 5% of 10% of 50% of 90% of
Country of origin # firms (€ bn) ers turnover turnover turnover turnover turnover turnover turnover turnover
Germany 48325 488.66 59.34 46.89 40.30 11.85 0.96 99.49 98.54 73.57 5.95
France 23691 171.73 67.30 41.16 33.04 9.02 1.39 93.58 95.11 49.22 9.71
United Kingdom 14976 71.46 28.33 22.52 19.27 8.07 1.51 97.60 93.40 65.70 19.00
Italy 4159 58.61 74.44 64.90 57.42 25.58 2.91 99.71 98.53 69.09 7.52
Hungary 6404 30.01 47.53 38.43 34.74 22.19 11.01 99.86 99.64 92.01 69.13
Norway 8125 16.07 39.22 17.98 14.45 5.19 1.26 98.51 97.42 70.27 28.57
N o t e : Germany, Italy, Hungary, the United Kingdom and France have large firms only; Norwegian data are exhaustive.
Figure 3 France having slightly more of both very small and very
Export Intensity: France versus Germany large exporters. In 2003 the picture is quite different,
a. 1998 with Germany outperforming France for middle-size
75 exporters by a fairly large margin. Whether this change
50 in distribution can explain the drastic differences in ex-
port performance of the two countries over the same
share of firms
S o u r c e : EFIM.
Table 4
Exporters and FDI-makers Exhibit Superior
product to only one market while 10% of firms export Performance
more than ten products to more than ten markets. Country of Employ- Value added Wage Capital Skill
origin ment premia premia intensity intensity
The bottom panel reports the shares of aggregate premia premia premia
exports due to firms exporting given numbers of prod-
Exporters premia:
ucts (rows) to given numbers of markets (columns).
2.99 1.02
The bipolar pattern is not there: firms exporting more Germany
(4.39) (0.06)
than ten products to more than ten markets account 2.24 2.68 1.09 1.49
France
for more than 75% of total exports. (0.47) (0.84) (1.12) (5.60)
United 1.01 1.29 1.15
Comparing the two panels then yields: Kingdom (0.92) (1.53) (1.39)
2.42 2.14 1.07 1.01 1.25
Italy
Fact 3 – Top exporters export many products to (2.06) (1.78) (1.06) (0.45) (1.04)
many locations. Firms exporting more than ten 5.31 13.53 1.44 0.79
Hungary
(2.95) (23.75) (1.63) (0.35)
products to more than ten markets account for
9.16 14.80 1.26 1.04
more than 75 % of total exports. Belgium
(13.42) (21.12) (1.15) (3.09)
6.11 7.95 1.08 1.01
To summarise, aggregate exports are determined Norway
(5.59) (7.48) (0.68) (0.23)
by a few top exporters that are relatively big and sup-
FDI- makers premia:
ply several foreign markets with several differentiated
13.19
products. This points to the existence of a process Germany
(2.86)
through which only firms that are large enough and 18.45 22.68 1.13 1.52
France
have a rich enough portfolio of products can withstand (7.14) (6.10) (0.90) (0.72)
international competition. We shall explore below the 16.45 24.65 1.53 1.03
Belgium
(6.82) (11.14) (1.20) (0.82)
characteristics that make exporters, and a fortiori top 8.28 11.00 1.34 0.87
Norway
exporters, different from other firms. We shall refer to (4.48) (5.41) (0.76) (0.13)
such differences as “exporters’ premia”. N o t e : The table shows premia of the considered variable as the ratio
of exporters over non-exporters (standard deviation ratio in brackets).
As to market coverage, most naturally the larger the France, Germany, Hungary, Italy and the United Kingdom have large
number of markets a firm serves, the larger their aver- firms only; Belgian and Norwegian data are exhaustive.
age distance from the firm’s country of origin. Table 3 S o u r c e : EFIM.
Table 5 presents, instead, two measures of produc-
tivity for French exporters. Revenue per worker is re-
$ENSITY
corded as “apparent labour productivity”. “Total factor
productivity” (TFP) refers to the estimated productivity
of all inputs taken together and is a measure of the
global efficiency of a firm.
The message conveyed by the two tables is clear:
in all countries and on all counts exporters are gen- ,ABOURæPRODUCTIVITY
erally better performers. The difference is particularly
B
pronounced for employment and value added. There
is, nonetheless, some variation across countries. For
France (2.4 and 2.6) and Italy (2.2 and 2.1) than for Bel-
gium (9.1 and 14.8) and Norway (6.1 and 7.9). This is
probably due to the fact that the French and the Ital-
ian datasets feature relatively large firms only, which
gives highly selected samples of non-exporters. The
wage premium is, instead, consistently smaller but still
4&0
exporters tend to pay wages that are 10-20% higher $OMESTIC
%XPORTERSæANDæ&$)
than non-exporters. %XPORTERS
The employment premium for German exporters is N o t e : Data for Belgium 2004.
in line with those of France and Italy. The United King- S o u r c e : EFIM.
Figure 5 Figure 6
The Rising Foreign Ownership of European Compared Performance in Labour Productivity –
Exporters Exports (France)
(share of foreign-owned exporters in %)
VALUEæADDEDæPERæWORKERYEARææEUROS
50
40
30
20
10
SWITCHæYEAR SWITCH SWITCH SWITCH
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 NONçSWITCHERS SWITCHERS
Belgium Italy Hungary United Kingdom
S o u r c e : EFIM. S o u r c e : EFIM.
For the three types of firm, each panel shows the tion of these figures over time. Hungary and the UK
share of firms (“density”) that attain each productivity are quite stable in having a very large share of foreign-
level. In other words, the panels depict the probability owned exporters, while foreign ownership is rising fast
of picking a firm with a certain productivity level when in Belgium and Italy.
the firm is randomly drawn from each type. The two
Hence, we have:
panels send the same message: a randomly drawn
FDI-maker is likely to be more productive than a ran- Fact 5 – Exporters are more likely to be foreign
domly drawn exporter, which in turn is likely to be more owned.
productive than a randomly drawn domestic firm. This
type of finding is not specific to Belgium, and has also Learning by Exporting and Investing Abroad?
been shown to exist for Italian exporters compared to Exporters are better than non-exporters over a
domestic Italian firms.5 broad spectrum of performance measures. An inter-
We have therefore established: esting issue is whether their superior performance
predates their access to export markets or, rather,
Fact 4 – FDI-makers perform better than exporters their performance improves as a result of their access
and exporters perform better than non-exporters. to export markets.
Exporters are generally bigger, more profitable,
This chicken-and-egg question is presented for
more capital intensive, more productive and pay
France and Norway in Figure 6 and Figure 7 respec-
higher wages than non-exporters. By the same
tively. The figures consider firms in the samples that
measures, FDI-makers perform better than export-
ers.
owned. The associated Figure 5 depicts the evolu- N o t e : United Kingdom, Italy and Hungary have large firms only, Bel-
gian data are exhaustive.
5
http://www.cepr.org/pubs/PolicyInsights/PolicyInsight8.pdf. S o u r c e : EFIM.
Figure 7 Figure 8
Compared Performance in Labour Productivity – Compared Performance Ratio in Labour
Exports (Norway) Productivity for Three Countries – Exports
2
VALUEæADDEDæPERæHOURæWORKEDææ./+
1
0.5
0
switch year switch+1 switch+2 switch+3 switch+4
SWITCHæYEAR SWITCH SWITCH SWITCH SWITCH Germany France Norway
NONçSWITCHERS SWITCHERS
S o u r c e : EFIM. S o u r c e : EFIM.
became exporters during the period of observation the switchers were already on a better trajectory, so
and that were observed for four years after switching gaining export status was simply the outcome of an
status (“switchers”). It then compares their behaviour already promising performance (“selection into export
with that of all other firms (“non-switchers”). In particu- status”). On the other hand, perhaps the switchers
lar, the comparison is made in terms of value added were no different from other firms before switching,
per worker a given number of years after the firms first
but gaining export status as a result of some tempo-
began exporting.
rary shock allowed them to learn from international
The two figures show that switchers do move along activity (“learning-by-exporting”). Data for Germany
steeper trajectories as they perform increasingly bet- are also available but only allow one to calculate the
ter than non-switchers. This is true no matter wheth- performance ratios of switchers over non-exporters.
er they already performed better in the switch year
We compute those ratios for the three countries and
(France) or not (Norway). Two very different stories
depict them in Figure 8.
are consistent with those findings. Since we do not
observe what happened before the switch, perhaps While the labour productivity of firms switching to
exporter status is generally greater than that of non-
exporters one year or more after switching, the pattern
Figure 9 over time is not clear. The advantage increases steeply
Compared Performance Ratio in Labour
for Norway but much less so for France and does not
Productivity – FDI (Norway)
show any clear trend in the case of Germany.
VALUEæADDEDæPERæHOURæWORKEDææ./+
Figure 10 the case of both exports and FDI, for ease of presenta-
The Margins of Adjustment of Aggregate Exports tion we shall initially focus on trade flows and deal with
Firm extensive margin Product extensive margin Quantity margin FDI later.
Aggregate data show that bilateral trade flows are
Number of
exporters positively affected by countries’ sizes and negatively
affected by trade impediments. As some trade impedi-
Aggregate ments increase with the distance between countries,
value of Number of
exports exported this result is reminiscent of Newton’s law of gravita-
Value products Quantity exported
exported per product
tional attraction, whence the name “gravity equation”.
per Exports per per exporter
exporter product per Through which channels does gravity determine bi-
exporter Export price per
product per
lateral trade flows? First of all, gravity may affect the
exporter number of exporters (“firm extensive margin”). Then,
it may affect the average exports per exporter (“firm
Firm intensive margin Product intensive margin Price margin
intensive margin”). It may also affect the number of
S o u r c e : EFIM.
products exported (“product extensive margin”), and
the average exports per firm of each product (“prod-
uct intensive margin”). Finally, gravity may affect ex-
Fact 6 – There is no clear evidence of firms per- port prices (“price margin”) and exported quantities
forming differently after accessing foreign markets. (“quantity margin”) in different ways. To handle this
While the performance of firms that start exporting complexity in a consistent way, we decompose the
is generally better than that of non-exporters one simple gravity equation into increasingly finer detail,
year or more after starting to export, the pattern over relying on firm-level information. The logic of this de-
time is not clear. The picture is even more blurred in composition is visualised in Figure 10.
the case of firms that start to invest abroad. Let us begin with the decomposition in terms of firm
The Margins of Exports and FDI extensive and intensive margins. In other words, we
ask: do spatial separation, differences in language,
This section breaks down aggregate exports and
currencies and so on hinder trade flows by limiting the
FDI into their fundamental drivers. It shows that the
entry of exporters (“firm extensive margin”) or rather
most important channel through which these drivers
by constraining the volumes exported by firms (“firm
affect aggregate flows is the “extensive margin”, i.e.
intensive margin”)?
the number of IFs.
The decomposition of exports into extensive and in-
The single most robust way to relate aggregate
tensive margins can be carried out in a similar fashion
trade and FDI flows to their fundamental drivers is the
for the French and Belgian data, which both provide
“gravity equation”. This relates the values of flows be-
near-exhaustive data for exports over a very compara-
tween two economies to their sizes and a variety of
ble set of years. Furthermore, we are able to compute
trade impediments.6 While this relationship works in
for both countries not only the average export value
6
per firm, but also the number of products exported,
The theoretical foundations of this empirical relationship have
emerged late in time compared with the vast number of empirical ap- the average quantity (in kilograms) and therefore the
plications of gravity. In the last ten years a wide range of theoretical unit value for each product.7
explanations for gravity have become available (see J . E . A n d e r-
s o n , E . v a n W i n c o o p : Trade Costs, in: Journal of Economic Lit- We start with the most simple decomposition exer-
erature, Vol. 42, 2004, pp. 691-751 for a survey), and researchers such
as Chaney, Helpman et al. and Melitz and Ottaviano have started to cise, which contains only distance as a trade impedi-
investigate the importance of firms’ heterogeneity for gravity. (Cf. T. ment. Figure 11 presents the results.8 The bar chart
C h a n e y : Distorted Gravity: The Intensive and Extensive Margins of
International Trade?, in: American Economic Review, 2006, forthcom- represents the contribution of firm extensive (“number
ing; E . H e l p m a n , M . J . M e l i t z , Y. R u b e n s t e i n : Estimating of exporters”) and intensive (“average exports”) mar-
Trade flows: Trading Partners and Trading Volumes, in: Quarterly Jour-
nal of Economics, 2007, forthcoming; M . M e l i t z ; G . O t t a v i a n o : 7
We can thus go even further than existing margin decomposition
Market Size, Trade, and Productivity, in: Review of Economic Studies, on US internal (R . H i l l b e r r y, D . H u m m e l s : Trade Responses
Vol. 75, 2008, pp. 295–316). On the empirical side, authors such as J . to Geographic Frictions: a Decomposition Using Micro-Data, in: Eu-
E a t o n , S . K o r t u m and F. K r a m a r z : An Anatomy of Interna- ropean Economic Review, 2005) or external (B . B e r n a r d et al., op.
tional Trade: Evidence from French Firms, in: Meeting Papers from So- cit.) data. Another early paper decomposing trade patterns into the
ciety for Economic Dynamics, No. 197, 2005; and A . B . B e r n a r d , extensive and intensive margins is J . E a t o n et al., op. cit., using
J . B . J e n s e n , S . J . R e d d i n g and P. K . S c h o t t : Firms in Inter- French data for the year 1986.
national Trade, in: Journal of Economic Perspectives, Vol. 21, No. 3,
8
2007, have investigated those issues for US firms and French firms. All coefficients are highly significant.
Figure 11 Figure 12
Gravity and Aggregate Exports – I Gravity and Aggregate Exports – II
1.5 2.5
2
1
1.05 1.5
0.93 1.05
1
0.5
0.93
0.5
0 0
GDP, im Dist. GDP, ex GDP, im Dist.
GDP, ex -0.5
0.5 -1
-0.86
0.86 -1.5
1 Number of exporters Number of products
Number of exporters overall effect Average export per product by firm Overall effect
Average Export
S o u r c e : EFIM. S o u r c e : EFIM.
gins to the overall effects (small diamonds) of three pediments. Sharing a language increases the number
gravity forces on bilateral exports: the size of the ex- of exporters and does not affect the average amount
porting country (“GDP, ex”), the size of the importing exported. GATT/WTO membership and colonial links
country (“GDP, im”) and distance (“Dist.”) increase the number of exporters and reduce the av-
The overall effects are extremely standard: close to erage amount exported. This evidence is compatible
one for GDPs and close to -0.9 for distance. In other with the notion that being a member of GATT/WTO
words, if country A is 10% larger than country B, then and having linguistic or colonial links tends to reduce
on average it attracts 10% more exports than B from the fixed costs of exporting rather than the variable
other countries. Analogously, country A exports on av- ones.
erage 10% more than B to other countries. Moreover, We have thus established:
if A is on average 10% further away from other coun-
tries than B, then it trades 9% less than B with those Fact 7 – The number of exporters matters the most.
countries. The change in the number of exporting firms ac-
counts for most of the negative impact of trade
More interestingly, the results of the decomposition
barriers and most of the positive impact of the im-
show that the reaction of the firm extensive margin of
porting country’s size on bilateral exports. The in-
trade to gravitational forces is much greater than that
crease in the number of exporting firms accounts
of the intensive margin. For instance, the decrease in
entirely for the positive impact of the exporting
the number of firms accounts for 75 % of the impact of
country’s size on bilateral exports.
distance on trade flows. In the same vein, the increase
in trade value associated with the increase in the im- Product Margins
porting country’s size comes mostly (60%) from the In datasets where the information is available, a fur-
increase in the number of exporters to the country in ther decomposition makes it possible to assess how
question. Note also that the entire effect of the export-
the number of products exported by firms varies with
ing country’s size on trade comes from the number of
different barriers to trade.
its exporting firms.9
Figure 12 displays the results of this new decom-
More detailed estimates also allow us to identify in-
position. The bar chart represents the contribution
teresting differences in the effects of different trade im-
of the firm extensive margin (“number of exporters”),
9
This is exactly what should be expected from most theoretical foun- the product extensive margin (“number of products”)
dations of the gravity equation and, in particular, from the ones with
differentiated products and imperfect competition, whether with het-
and the product intensive margin (“average exports
erogenous firms (T. C h a n e y, op. cit.; E . H e l p m a n et al., op. cit.; per product by firm”) to the overall effects (small dia-
M . M e l i t z , G . O t t a v i a n o , op. cit.) or not. (S. R e d d i n g , A . Ve -
n a b l e s : Economic Geography and International Inequality, in: Jour-
monds) of three gravity forces on bilateral exports.
nal of International Economics, Vol. 62, 2004, pp. 53-82). Strikingly, the results point to an extreme parallelism
144 Intereconomics, May/June 2008
COMPETITIVENESS
the corresponding parts of the bars, the effect of these
three factors on exports is mitigated by their effect on
average exports per product by firm. Indeed, the av-
erage exports per product by firm fall with GDP and
rise with distance. In particular, a 10% increase in the
GDP of the exporting country leads to an increase of
roughly 10% in both the number of exporters and the
number of products exported as well as a decrease of .UMBERæOFæEXPORTERS .UMBERæOFæPRODUCTS
!VERAGEæQUANTITYæPERæPRODUCTæBYæFIRM /VERALLæEFFECT
roughly 10% in firms’ average export per product. A !VERAGEæPRICEæPERæPR ODUCTæBYæFIRM
10% increase in bilateral distance leads to a 6% fall S o u r c e : EFIM.
in the first two variables and to a 4% increase in the
third.10 lated to “quality sorting” of firms over different export
markets. The former refers to the fact that only the
These findings establish:
most productive firms from a certain country manage
Fact 8 – The number of exported products matters to export to distant or small foreign markets. This oc-
too. Larger countries have more exporters, export curs because only those firms are able to quote low
more products and their exporters have smaller av- enough prices but still succeed in exporting large
erage exports per product. An increase in bilateral enough quantities to at least break even. Nearer or
trade barriers reduces the positive effects of coun- larger markets attract many more exporting firms, and
try size on the number of exporters and products. It the proportion of high cost – high price – low quantity
also reduces the negative effect of country size on exporters is larger.12 Since the product intensive mar-
exporters’ average exports per product. gin only considers the average shipment value, such
a composition effect may explain why the effects of
The results on the product intensive margin are par-
GDP are negative and those of distance are positive.
ticularly interesting. They imply that the indications of
the (net) impacts of GDPs and distance on the firm in- Alternatively, the puzzling signs of the effects may
tensive margin highlighted in Figure 11 are attributable have to do with the quality or price/weight ratio ex-
to their impact on the total number of exported prod- ported to different markets. If firms differ in the qual-
ucts, which is far greater than the impact on average ity of the product exported (or have different qualities
exports per product.11 in their portfolio of products), it may be observed that
only the high quality varieties are exported to distant or
We can thus write:
small markets, while low quality products can only be
Fact 9 – Firms’ average exports per product matter exported to nearer or large markets.13 Distinguishing
less. The changes in the number of exporting firms between the two alternative explanations is a complex
and in the number of exported products accounts issue, but the average price of shipments can be used
entirely for the negative impact of higher trade bar- to shed some light on it.
riers and the positive impact of larger countries’ Price and Quantity Margins
size on bilateral exports.
We now turn to the last decomposition, which al-
The finding that the “product intensive margin” falls lows us to distinguish between the gravity effects on
with GDP and increases with distance is puzzling at average quantity and on average price.
first sight. Two hypotheses can be proposed to explain A final decomposition of the average exports per
it, one related to “efficiency sorting” and another re- product by firm (product intensive margin) into aver-
10 12
These findings are very similar to the ones by A . B . B e r n a r d et See M . M e l i t z , G . O t t a v i a n o , op. cit., for a theoretical formali-
al., op. cit., and R . H i l l b e r r y, D . H u m m e l s op. cit., for external sation of this idea.
and internal US trade flows, respectively. 13
A . B . B e r n a r d et al., op. cit., conjecture that this second expla-
11
This finding parallels the one by A . B . B e r n a r d et al., op. cit., for nation might be the relevant one to explain their result, but do not in-
US exporters. vestigate it further.
age quantity per product by firm and average price per Further analysis of French data reveals that former
product by firm can be carried out using information colonial ties and sharing a common language have a
on the value and quantity of shipments measured at positive impact on trade flows along both the intensive
product level. The results of this final decomposition and the extensive margins. Moreover, in markets that
are reported in Figure 13. are easier to access, such as those of former colonies
and francophone countries, French exporters are more
The bar chart in Figure 13 represents the contri- numerous and on average less efficient, which drives
bution of the firm extensive margin (“number of ex- down the average quantity exported. As colonial ties
porters”), the product extensive margin (“number of and a common language are not directly related to
products”), the quantity margin (“average quantity distance, that suggests that such variables proxy for
per product by firm”) and the price margin (“aver- lower fixed costs of exporting.
age price per product by firm”) to the overall effects
(small diamonds) of three gravity forces on bilateral Hence, we can highlight:
exports.
Fact 11 – Historical ties and common language
The chart shows some support for both “efficiency matter. Historical ties such as former colonial links
sorting” and “quality sorting”. The former implies that and a common language foster exports, making it
firms managing to export to smaller or more distant easier for less efficient firms to export.
markets are on average more productive and therefore Finally, the relationship between market size and
have on average higher volumes of sales. Figure 13 average prices is not as clear as the other three rela-
shows that the average quantity exported decreases tionships. This is not unexpected, since this average
with GDP and increases with distance, pointing to the price is a mixed bag of all sorts of underlying product
presence of less efficient exporting firms in larger or prices.
closer markets. The dark areas report the results for
the average unit price of exports. Average prices tend The Margins of FDI
to increase with distance from the exporting country,
The gravity model has been primarily devoted to the
which is consistent with “quality sorting”, as long as
study of trade flows, but more recently a fair amount of
higher quality varieties are the only ones able to reach
research has used the same variables to explain pat-
distant markets. However, such a mechanism would
terns of bilateral FDI flows or stocks.15 The equilibrium
certainly predict a negative effect of GDP. Hence, over-
equation for bilateral capital flows closely resembles
all “quality sorting” seems to be a weaker explanation
the gravity relation for bilateral trade flows. Nonethe-
of the aggregate observed behaviour of the product
less, the interpretation of the coefficients is sometimes
intensive margin.14
very different. Most importantly, in the case of trade
We have therefore: flows the negative coefficient on distance captures the
frictions due to trade costs (including freight costs),
Fact 10 – Prices and quantities defy gravity. The while in the case of FDI flows the same coefficient
average quantity exported by firms and the average captures the frictions due to information and transac-
export price per product are, respectively, smaller tion costs associated with the acquisition or installa-
and larger in larger countries. A reduction in trade tion of new capital abroad.
barriers leads to a fall in both of them.
As in the case of bilateral exports, the decomposi-
tion of the margins can be used to highlight the chan-
14
One must be cautious in interpreting these results since validating
nels through which gravity forces affect the sales of
the “quality sorting” hypothesis would imply running the analysis at
the firm level and measuring quality more directly. More generally, the foreign affiliates. In Figure 14 each bar chart repre-
average price is a mixed bag of all sorts of underlying product prices,
and therefore trade composition effects are likely to blur any story con-
15
cerning efficiency or quality sorting at the industry or even firm level. For example, K. H e a d , J. R i e s : FDI as an Outcome of the Market
Those sorting effects could only be properly uncovered through care- for Corporate Control: Theory and Evidence, in: Journal of Interna-
ful firm or industry level analysis, which goes well beyond the scope tional Economics, Vol. 74, Issue 1, 2008, pp. 2-20 have recently de-
of the present descriptive analysis. See R . B a l d w i n , J . H a r r i - veloped a model of FDI where heterogeneous investors bid to obtain
g a n : Zeros, Quality and Space: Trade Theory and Trade Evidence, in: control rights to existing overseas assets. The equilibrium equation for
NBER Working Paper No. 13214, 2007; and M . C r o z e t , K . H e a d , bilateral capital flows closely resembles the type of trade flow gravity
T. M a y e r : Quality sorting and trade: Firm-level evidence for French equation derived with heterogeneous exporters. In the same spirit, A.
wine, in: mimeo, CEPII, 2007 for more detailed hypotheses on this is- H i j z e n , H. G ö r g , M. M a n c h i n : Cross-border mergers and acqui-
sue. Deeper investigation is also needed to shed light on additional sitions and the role of trade costs, in: European Economic Review,
issues such as the opposite effects of regional trade agreements (RTA) 2007, forthcoming, investigate the role of trade costs in explaining the
on average export price and average export quantity. increase in the number of cross-border M&As.
tionalised firms”). The analysis of firm-level evidence and FDI at home. Trade missions do not generate
reveals some new facts that are simply unobservable trade.18
at the aggregate level:
• Fourth, nurture the superstars of the future. Govern-
ments should provide the conditions for tomorrow’s
• IFs are superstars. They are rare and their distribu-
superstars to emerge by allowing small exporters
tion is highly skewed, as a handful of firms accounts
and multinationals to grow.
for most aggregate international activity.
• Fifth, keep up the fight against small-trade costs.
• IFs belong to an exclusive club. They are different Small (fixed) costs of internationalisation matter be-
from other firms. They are bigger, generate higher cause they reduce the number of exporters and mul-
added value, pay higher wages, employ more capital tinationals.
per worker and more skilled workers and have higher
productivity. • Sixth, assess the export and FDI potential of your
industries. Some industries are more likely than oth-
• The pattern of aggregate exports, imports and for- ers to react to shocks through adjustment in the
eign direct investment is driven by the changes in numbers of exporters and FDI-makers. Hence, they
two “margins”. The “intensive margin” refers to aver- have greater unexploited export and FDI potential.
age exports, imports and FDI per firm. The “exten- These are industries characterised by a larger pres-
sive margin” refers to the number of firms actually ence of small, low-productivity firms. As such, they
are also more likely to react to import competition
involved in those international activities.
through the exit of the worst-performing firms and
• The “extensive margin” is much more important, therefore also have greater unexploited productivity
as the reaction of aggregate trade and FDI flows to gains from selection.
country fundamentals takes place mostly through Our findings also leave some questions open. We
that margin. This is impossible to see without firm- prioritise six of them for future investigation. If firms
level data and thus has not been seen so far. have to be large to be competitive in international mar-
kets, what is the significance of the size of the internal
We stress six policy implications. market? If superstars dominate international markets,
is there any room for global SMEs? What precisely
• First, promote intra-industry competition. The does the dominance of the extensive over the inten-
opening up of trade and FDI triggers a selection sive margin imply for policy intervention designed to
process in which the most productive firms replace promote the internationalisation of European firms?
the least productive ones within sectors. This is Do firms improve their performance when exposed
good for productivity, GDP and wages, even when to international competition? Is the fragmentation of
it does not lead to sectoral specialisation. Moreo- production processes across countries a way through
ver, precisely because winners and losers belong to which firms become more competitive in international
the same sector, the benefits of selection are likely markets? Is the limited internationalisation of European
to be associated with limited social costs of adjust- firms eroding political support for the single market?
ment.
Answering these questions requires quality data
• Second, increase the number of exporters and mul- at the firm level to be representative and comparable
tinationals. What matters most for a country’s trade across European countries. Currently, however, the
and FDI performance is, first of all, how many of its overlap among the different national datasets in terms
firms engage in exports and FDI. So governments of several key variables is far from complete at the tar-
geted level of disaggregation. In this paper we select
should focus on policies that broaden the export
different countries depending on the specific issues
base.
addressed. This is clearly a second-best approach
• Third, forget the incumbent superstars. If the aim is that is nevertheless enough to highlight the benefits
that would come from the creation of a harmonised
to broaden the export base, governments should
European dataset.
not focus on policies that favour existing superstar
exporters and multinationals. Instead, heads of gov- 18
See J. H e a d , K. R i e s : “Do trade missions increase trade?”, Uni-
ernment should work on lowering barriers to exports versity of British Columbia, mimeo, 2007.