Trading with Latin America and the Caribbean:
an economic geography of coal (1890-1930)1
Carreras-Marín, Anna
(Universitat de Barcelona, [email protected])
Badia-Miró, Marc.
(Universitat de Barcelona, [email protected])
The aim of this paper consists in understanding the dynamics of coal trade between Latin America
and its main trade partners, i.e. USA, Great Britain and Germany. That's not only a sectorial
approach as long as coal has been used as a modernization indicator for the Latin American
countries, taking into consideration that oil wasn't still enough important. Energy imports have
determined the possibilities of each Latin American country in its development stage. We focus on
that question emphasizing the channels of that supply. As a result we conclude that the economic
link with the USA and UK has accounted for much of the modernization of that countries. We find
that this is specially clear at the end of the period, but it has started well before the First World War,
at least for the majority of LA&C countries. We argue it through a market share analysis, based on
cluster methodologies, which allows as to classify the LA&C countries in separately groups. We
have also developed a gravity model to explained the factors of coal trade, as an indicator of
modernization chances.
1 This article is the result of research at an early stage of a project entitled ‘Imports and economic modernization in
Latin America 1890-1960’, financed by the Spanish Ministry of Education (Project No: BEC2003-00190 MCYT).
We are obliged to the rest of the team members for their encouragement and help: A.Carreras, M. Folchi, J. Jofré, F.
Notten, C. Román, M. Rubio, X.Tafunell, C.Yañez, S. Kuntz and A.Hofman. This investigation has been also partly
supported by the project SEJ2005 – 02498 / ECON (MEC, Spain), supervised by J. Catalan. The authors gratefully
acknowledge the research assistantship of F. Notten and C. Román. The usual disclaimers apply.
1 INTRODUCTION
Foreign trade had deserved much of the attention on the history of Latin America and the Caribbean
(LA&C as follows) for the period before the First World War2. We have developed an economic
geography approach to that question, focusing on the suppliers as well as developing a model to
explain its causes. Explaining the coal trade channels we pretend to approach the modernization
chances of LA&C before and after the First World War.
The paper is organized as follows. The first section is an overlooking of total LA&C trade with UK
and USA. We have emphasized the differences between each supplier following a geographical
approach. The second section is quite similar but focusing only the coal trade. In that case, we have
included Germany, because its share over LA&C coal imports were of some importance at the end
of the period.
In the third section we attend the relationship between the USA/UK trade dependency and the First
World War. We have used a cluster methodology to identify statistically common share patterns
among LA&C countries. We have included also a historical perspective, analysing the change of
such structures over time. We have find some interesting results. British traditional markets, i.e.
Argentina, Brazil and Uruguay, were relative stable in its British coal imports, even after the First
World War. On the other hand, for a quite big sample of small countries the war meant a trade
partner substitution, from UK to USA. But for the majority of LA&C countries, USA was the main
coal supplier well before the war. The few coal producers present a differentiated pattern in which
import substitution is the main feature. For another countries we have also identified the impact of
the opening of the Panama Canal in 1914, which changed a lot the distances in the continent.
The last section presents a gravity model to explain coal trade among LA&C countries and UK /
USA. As a result we find that trade opening but also trade partner shares played and important role
in the modernization chances of these countries. The more total imports of a country the more coal
imports, allowing for an increasing and earlier modernization. The more British dependency the
more coal imports, but the more USA dependency the less coal imports. The explanation is that
USA promoted an energy transition from coal to oil, and due to that it appears with a negative sign
for the coal trade.
2 See as an example, Bulmer-Thomas (1998), chapter 3, 4, and 5.
2 LATIN AMERICAN AND CARIBBEAN IMPORTATION FROM
USA AND UK
The insertion of LA&C in the world-wide markets in the First Globalization follows different
results based on the country that we are analysing. Anyway, as a whole, we can speak of a weak
insertion in the world-wide markets by the own instability of these, and the concentration of the
exports in few products3. In spite of it, imports become an essential indicator for the consumption
and investment for these countries because most of the manufactured goods and energy was
provided by foreign suppliers, when only few countries could supplied by itself4.
We have demonstrated in a previous work the high quality of LA&C Foreign Trade Statistics, when
compared with that from the suppliers5. But if we want to include as much LA&C countries as
possible we have no alternative but to use the statistics from its suppliers. We don't have Foreign
Trade Statistics for all these countries, neither for all the period (1890-1930). Geographical and
historical coverage has determined the use of USA's and UK's sources, complemented for the coal
data from Germany.
The analysis of the total exports of the United Kingdom and the United States of America to LA&C
countries is a proxy for total exports of developed countries to the region. Total exports per capita
of USA and UK (as follows G2) to LA&C show the existence of different groups of countries. We
observe countries that throughout the period remain in privilege places (Argentina, Uruguay, Cuba
and Chile), but others with higher levels of imports in 1890 had felt in 1913 to low levels in the
context of the region (the case of Haiti or Peru), effect of the lottery of the natural resources6 (see
table 1).
An interesting part of our analysis is the number of countries included. The big ones have been
broadly treated in many works but for the smallest, there is still much to know. Big countries have
been used to explained the whole LA&C history. As we can see in table 1, including these small
countries offers a quite different story.
Among the big LA&C countries we find different levels of openess. Some of them were very
opened to the world-wide markets, like Argentina, Cuba and Chile, and others much more closed in
relative terms, like Mexico and Brazil (lower exports per capita). Among the small countries,
Uruguay and Costa Rica had remained very opened along the period 7 as some of the territories and
the colonies (extremely linked with their metropolis).
The Centre American and the Caribbean countries started with a low development over the rest of
Latin America, but the First Globalization seems to had been good for them. They gained some
positions in a context of growth of exports per inhabitant in the region. The First World War, the
collapse of the globalization, and the postwar recovery represented a better behaviour for the big
countries of the South Cone. As a consequence at the end of the period things seems to had came
back to the beginning. Small countries felt down to its previous position after the end of
globalization.
3 Bulmer-Thomas,1998, Thorp,1998, Hofman A,2000 and Carreras, A.,A.A. Hofman,X.Tafunell y C.Yáñez,2003 p.
11.
4 Yañez C, Rubio M & Carreras A,2006 for energy consumption.
5 Badia-Miró and Carreras-Marin (2005)
6 Bulmer-Thomas,1998
7 Notten (2005)
Table 1: Ranking of exports pc from UK and USA to LA&C, ($ 1913)
1890
Uruguay
Argentina
Peru
Cuba
Chile
Haiti
Costa Rica
Venezuela
Nicaragua
Brazil
Colombia
R. Dominicana
Mexico
Ecuador
Honduras
Salvador
Guatemala
Paraguay
Bolivia
22116.7
17429.0
11644.2
9217.6
8312.0
7990.7
5501.6
4102.9
3997.2
3989.9
2873.1
2337.1
2197.3
2145.2
1728.6
1508.6
1172.8
147.1
9.5
1913
Cuba
Argentina
Uruguay
Chile
Paraguay
Costa Rica
Salvador
Nicaragua
Panama
Honduras
Guatemala
Mexico
Brazil
Haiti
Ecuador
R. Dominicana
Peru
Colombia
Venezuela
Bolivia
1925
33718.5Panama
21318.8Cuba
15835.4Argentina
13085.5Uruguay
12758.6Costa Rica
12246.0R. Dominicana
8767.8Chile
7180.7Honduras
6627.7Nicaragua
6538.9Venezuela
4577.1Mexico
4370.5Salvador
4367.3Colombia
4100.6Guatemala
3138.4Haiti
3060.1Peru
3057.6Ecuador
3011.5Brazil
1510.8Bolivia
1343.2Paraguay
43008.6
40175.5
18173.6
15235.6
12403.1
11001.2
10959.5
9123.5
8494.6
8208.0
6828.3
6136.0
5833.2
5162.9
4551.4
4310.6
3836.0
3561.0
2193.4
1533.0
117797.1Bermuda
50261.8Brit. Honduras
17782.5British Guiana
16374.2French Guiana
133375.4
41811.0
18606.5
5863.3
Colonies and Territories
Bermuda
Brit. Honduras
British Guiana
80695.9
32343.5
26755.9
Bermuda
Brit. Honduras
British Guiana
French Guiana
If we focus on the trade between the main suppliers and its partners, we observe two important
facts: the existence of well defined geographical trade areas, and the different concentration of the
exports according to the USA or the UK. First, we consider the percentage that represents each one
of the exporters on the importer country (see table 2). We want to measure the dependency of a
country to each supplier. Second, we observe the trade concentration of the suppliers in a certain
geographic zone (see table 3 and 4).
Table 2: Percentage of United Kingdom exports as a part of Total G2 exports, over importer countries.
1890
Paraguay
Argentina
Chile
Peru
Uruguay
Brazil
Colombia
Ecuador
Venezuela
Mexico
Haiti
1913
100.0%S. Domingo
83.1%Venezuela
82.5%Panama
79.4%Uruguay
75.6%Argentina
75.3%Bolivia
68.8%Chile
66.4%Brazil
50.3%Peru
42.2%Colombia
33.5%Ecuador
Bermuda
Guatemala
Nicaragua
Costa Rica
Salvador
1925
100.0%S. Domingo
100.0%Paraguay
92.2%Argentina
71.2%Brazil
67.6%Chile
65.1%Uruguay
64.5%Peru
58.7%Bolivia
55.5%Ecuador
52.7%Venezuela
44.2%Colombia
35.1%Salvador
31.4%Guatemala
28.8%Costa Rica
24.7%Honduras
17.5%Nicaragua
100.0%
51.1%
48.6%
47.1%
42.6%
41.9%
33.3%
33.1%
32.8%
32.3%
31.1%
24.8%
21.7%
21.5%
19.5%
13.4%
Mexico
Honduras
Cuba
Paraguay
Haiti
R. Dominican
16.7%Haiti
16.3%Mexico
13.3%Panama
11.5%Cuba
10.9%R. Dominicana
0.0%
13.0%
9.5%
6.2%
6.0%
0.0%
Colonies and Territories
Bermuda
British G.
Dutch WI
British WI
British Hond.
French WI
Danish WI
Dutch G.
French G.
100.0%British G.
68.4%British WI
62.4%Dutc G.
61.2%British Hond.
58.5%Danish WI
45.5%French G.
41.5%Dutch WI
40.6%Frenc WI
15.1%Other Br. WI
66.2%British WI
47.1%British G.
32.2%Bermuda
28.9%Dutch G.
24.2%British Hond.
20.7%Dutch WI
19.3%French G.
17.9%French WI
0.0%
100.0%
76.6%
32.4%
31.5%
28.7%
23.4%
4.2%
2.2%
Although the countries with higher percentages remained along the time, British shares clearly fell
down. The values with higher percentage at the beginning were those of the South Cone, with
Argentina, Chile, Peru and Uruguay at the top. The First Globalization and the insertion of these
countries in the world-wide markets diversified its trade, being the imports of United Kingdom still
the most important, but with smaller percentage (beginning of the expansion of the USA exports in
the zone)8. At the same time, the important weight that had the foreign trade of the USA in Central
America and the Caribbean, was reinforced. The impact of the First World War, joint with the
opening of the Panama Canal, had an important effect on the diminishing total weight of the British
exports.
The British percentage of the top countries (with the exception of Paraguay) decreased; the USA
became the clear dominator of the region, as the United Kingdom did before. In spite of that,
differences between zones were remarkable. Meanwhile in the South Cone the British exports
surpass the third of the total of G2, in Central America and the Caribbean, that percentage was
reduced to less of 20% in most of them, even less in cases like Panama, Mexico or Cuba. That is not
the pattern followed by the coal trade, product that maintained, after First World War, a structure in
exports quite similar in both periods, as we will see next.
Table 3: Percentage of US exports by destination.
1890
Cuba
Mexico
Brazil
Argentina
Haiti
Venezuela
Chile
Uruguay
Colombia
Peru
Guatemala
Nicaragua
Costa Rica
R. Dominicana
Salvador
Ecuador
Honduras
8 Bethell ().
1913
14.8%Cuba
14.8%Mexico
13.9%Argentina
9.7%Brazil
5.9%Chile
4.6%Salvador
3.8%Colombia
3.7%Paraguay
2.9%Haiti
1.7%Peru
1.5%Uruguay
1.5%Guatemala
1.3%Costa Rica
1.1%Honduras
1.0%Nicaragua
0.8%Ecuador
0.6%R. Dominicana
1925
20.6%Cuba
15.8%Argentina
15.4%Mexico
12.4%Brazil
4.7%Colombia
2.2%Chile
2.2%Panama
2.1%Venezuela
1.9%Peru
1.7%Uruguay
1.7%R. Dominicana
1.1%Haiti
1.0%Honduras
0.9%Guatemala
0.9%Salvador
0.7%Nicaragua
0.7%Ecuador
23.1%
17.3%
16.8%
10.2%
4.8%
4.6%
3.3%
2.9%
2.7%
2.5%
2.1%
1.6%
1.1%
1.1%
1.1%
0.9%
0.8%
Bolivia
0.0%Bolivia
Panama
0.3%Costa Rica
0.1%Bolivia
Paraguay
0.8%
0.6%
0.1%
7.2%Dutch WI
3.7%Bermuda
0.5%French WI
0.5%British H.
0.4%British G.
0.4%Dutch G.
0.3%French G.
0.3%
0.2%
0.1%
0.5%
0.4%
0.3%
0.3%
0.2%
0.1%
0.0%
Colonies and Territories
British WI
British G.
French WI
Danish WI
Dutch WI
British H.
Dutch G.
French G.
9.4%Oth. British WI
2.3%British WI
2.1%British G.
0.9%French WI
0.7%Bermuda
0.4%British H.
0.3%Dutch WI
0.2%Danish WI
Dutch G.
French G.
The geographic distribution of British exports was much more concentrated than the North
American exports. In the British zone, the three main destinies concentrated more than 65% of total
exports (see table 4). In the USA's zone, the three main destinies didn't surpass, or they did it
slightly, 50% of the total (see table 3). Certainly, the big countries like Argentina, Brazil or Mexico
appeared to be the main destinies for both exporter countries, as expected for its size, but in both
cases appeared other countries: Cuba for the exports of the USA, and Chile and Uruguay for the
British exports.
Simultaneously, it is remarkable the higher British concentration over the North-American one (the
biggest British importer almost doubles USA one in every year, in percentages). The tendency for
the British exports shows an increase of the concentration in the First Globalization, and a decrease
in the later period. The same happened for the USA, for which a greater diversification is even
observed all over the period. The only country with certain relevance in the British exports in the
Caribbean zone, was Cuba, due to the importance of the country in all over Latin America9. The
imports of the USA arrived at the smallest countries of the South Cone, like Peru, Chile or Uruguay
already in 1913, staining, and even being increased later.
Table 4: Percentage of UK exports by destination.
1890
Argentina
Brazil
Chile
Uruguay
Mexico
Colombia
Peru
Venezuela
Haiti
Ecuador
Bermuda
Paraguay
1913
26.8%Argentina
23.8%Brazil
10.0%Chile
6.5%Uruguay
6.1%Mexico
3.6%Cuba
3.6%Colombia
2.6%Peru
1.7%Venezuela
0.9%Panama
0.7%Ecuador
0.0%Bolivia
Guatemala
Salvador
Nicaragua
Costa Rica
Paraguay
Haiti
Bermuda
St. Domingo
Honduras
9 Yañez et al. (2006). Rubio and Folchi (2005).
1925
38.3%Argentina
21.1%Brazil
10.2%Chile
4.9%Colombia
3.8%Uruguay
3.7%Mexico
2.9%Cuba
2.5%Venezuela
1.4%Peru
0.8%Ecuador
0.7%Salvador
0.6%Guatemala
0.6%Bolivia
0.6%Honduras
0.4%Haiti
0.4%Panama
0.3%Costa Rica
0.3%Bermuda
0.3%St. Domingo
0.3%Nicaragua
0.2%Paraguay
36.4%
20.2%
7.5%
4.8%
4.0%
3.9%
3.3%
3.1%
3.0%
0.9%
0.8%
0.7%
0.7%
0.6%
0.5%
0.5%
0.5%
0.4%
0.3%
0.3%
0.2%
Colonies and Territories
British WI
British G.
French WI
Dutch WI
Danish WI
British H.
Dutch G.
French G.
8.4%British WI
2.9%British G.
1.0%British H.
0.7%French WI
0.4%Dutch G.
0.3%Danish WI
0.1%Dutch WI
0.0%French G.
4.0%British WI
1.2%British G.
0.2%Dutch WI
0.1%British H.
0.1%Dutch G.
0.1%French WI
0.1%French G.
0.0%
5.0%
1.7%
0.3%
0.2%
0.1%
0.0%
0.0%
3 THE COAL SUPPLIERS: USA, UK AND GERMANY
Coal importation in LA&C countries began at the end of XIXth century. The almost absence of that
resource in the region made importation the only way to achieve it. In spite of that, some coal was
produced in Chile and Mexico since 1890, in Peru since 1900, in Brazil since 1912 and in
Venezuela in 1913. Anyway the LA&C coal production had a clear secondary role over the whole
region (see table 2). It accounted for only 15-29% of the total consumption.
That percentages are obviously greater for the few coal producers, but even in that cases, shares
vary a lot. Chile was the only one to export coal, mainly to Bolivia. Chile managed to produced as a
maximum of 98% of its coal consumption in 1900. Peru reached its maximum at the end of the
period, with a 87,7%. Mexico achieved the 90% only in the twenties, meanwhile for the period
before the First World War its share was quite low, under 50%. The Brazilian post-war coal
production hardly represented a 14-17% of its consumption. Venezuela was even worst, around 2%.
Dependency of coal importation was a common feature and it marked the chances of
modernization.
Table 2: Coal production as a share of apparent consumption in LA&C
1890
1900
1913
1925
1929
TOTAL
14,8%
23,2%
15,6%
28,7%
26,6%
Chile
73,0%
98,4%
58,4%
87,2%
96,5%
Peru
0,0%
73,5%
80,1%
75,2%
87,5%
Brazil
0,0%
0,0%
0,6%
16,8%
14,1%
Venezuela
0,0%
0,0%
2,0%
2,1%
1,6%
Mexico
46,0%
27,9%
47,8%
92,4%
93,1%
Source: Yañez C, Rubio M & Carreras A,2006
LA&C had to import coal from abroad, let's see who were the suppliers. In the international
markets, there were only three big coal exporters: USA, UK and Germany. Meanwhile German coal
was really marginal for the region, the huge amount came from UK (see graph 1). British coal
accounted for more than 50% during all the period, not considering the war years exception. But
that British high figures refer exclusively to a minority of countries: Argentina, Uruguay, Brazil and
the British colonies. For the other LA&C countries the big supplier was USA, even before the First
World War.
100%
9,000,000
90%
8,000,000
80%
7,000,000
70%
6,000,000
60%
5,000,000
IMP ALE %
4,000,000
40%
3,000,000
30%
2,000,000
20%
1,000,000
1929
1930
1927
1928
1925
1926
1923
1924
1921
1922
1919
1920
1917
1918
1915
1916
1913
1914
1911
1912
1910
1909
1907
1908
1905
1906
1903
1904
1901
1902
1899
1900
1897
1898
1895
1896
1893
1894
1891
1892
1890
10%
0%
IMP USA %
IMP UK %
50%
0
Illustration 1: Share of coal exports by country of origin, 1890 - 1930. Source: Yañez et al. (2006).
Total IMP G3
4
TRADE DEPENDENCY AND THE FIRST WORLD WAR
The First Word War had a big impact over total coal imports of LA&C, in terms of changing from
an increasing tendency to stagnation in the 20s. After the War, UK managed to almost recovered its
market shares, but USA and Germany had increased significantly. But if that's true for all the
countries as a whole, things change a lot for each one in particular. Trying to find common patterns
we have studied the share of each supplier, country by country, and its variation over time. We have
used a cluster methodology to identify such patterns, in which similarities are defined by statistical
criteria10. War years have been excluded of the sample, because we are interested in testing
structural changes before and after it.
A first group of countries show a clear British dependency in the two periods. Argentina, Brazil and
Uruguay imported around 90% from UK before 1914, and only a bit less in the 20s (see graph 2). A
common feature of these three countries were that they were not coal producers. Although it is an
exceptional situation, we can also observe the disappearance of UK during the war years. It is well
known the fact that USA invaded LA&C markets at that time, on the one hand because of absence
of European competitors, on the other because of increasing transport costs. But what is more
unexpected is the relative British recover.
120%
100%
80%
60%
Argentina
Brazil
Uruguay
40%
20%
añ
o
18
90
18
91
18
92
18
93
18
94
18
95
18
96
18
97
18
98
18
99
19
00
19
01
19
02
19
03
19
04
19
05
19
06
19
07
19
08
19
09
19
10
19
11
19
12
19
13
19
14
19
15
19
16
19
17
19
18
19
19
19
20
19
21
19
22
19
23
19
24
19
25
19
26
19
27
19
28
19
29
0%
Illustration 2: British share over coal consumption in Argentina, Brazil and Uruguay.
A second group of countries, started the period being British dependants and finished it being USA
dependants. The war shock has been lasting for this sample of LA&C countries, although for some
of them that's a fact that began even before 1914. Their small size explains the higher instability of
the results, compared to the big countries of the previous group. It is a bit surprising to find the
British colonies in that group, for which we would said a priori that UK will be able to maintain its
predominance. On the contrary, it seems that geographical proximity has accounted much than
political ties.
10 The cluster analysis is a methodology to group different objects which have similar patterns related to some chosen
variables. This allows us to find data structures without previous explanations or interpretations. Saber (1984).
Chapter 7. Cluster Analysis. p. 347 – 394.
120.00%
100.00%
80.00%
Bermuda
British Guiana
British Honduras
British West Indies
Costa Rica
Dutch Guiana
Haiti
60.00%
40.00%
20.00%
19
30
19
28
19
26
19
24
19
22
19
20
19
18
19
16
19
14
19
12
19
10
19
08
19
06
19
04
19
02
19
00
18
98
18
96
18
94
18
92
18
90
0.00%
Illustration 3: British share over coal consumption
A third group is defined by its high USA dependency since the beginning (see graph 4). They all
were non coal producers. USA were for most of them the only coal supplier, having a 100% of their
imports for almost all the years. Such countries were Barbados, Dominican Republic, French
Guiana, Honduras, Jamaica, Nicaragua, Trinidad and Tobago, Panama and Paraguay. The First
World War had no influence on them, as long as the USA predominance had been a previous
phenomenon.
120.00%
100.00%
Barbados
80.00%
Cuba
Danish West Indies
Dominican Republic
French Guiana
Honduras
60.00%
Jamaica
Nicaragua
Panamá
Paraguay
Trinidad & Tobago
40.00%
Venezuela
20.00%
3
0
19
2
8
19
2
6
19
2
4
19
2
2
19
2
0
19
19
18
19
16
19
14
19
12
10
19
19
0
8
19
0
6
0
4
19
0
2
19
19
0
0
8
18
9
6
18
9
18
9
4
18
9
2
18
9
0
0.00%
Illustration 4: USA share over coal consumption
For a forth group of LA&C countries, USA, being the main supplier, was not so stable predominant
as in the precedent group (see graph 5). The First World War appears here to be joint with the
opening of the Panama Canal in 1914. That engineering enormous infrastructure allowed to shorten
crucially the distances, as ships no longer had to travel the long and treacherous route via the Drake
Passage and Cape Horn. For example, shipping British coal to Ecuador meant around 11000 km
before 1914 and only 5000 km from the canal on.
1.2
1
0.8
Colombia
Ecuador
El Salvador
French West Indies
Guatemala
0.6
0.4
1929
1930
1927
1928
1925
1926
1923
1924
1921
1922
1919
1920
1917
1918
1915
1916
1913
1914
1911
1912
1910
1909
1907
1908
1905
1906
1903
1904
1901
1902
1899
1900
1897
1898
1895
1896
1893
1894
1891
1892
0
1890
0.2
Illustration 5: USA share over coal consumption
The firth group is composed by the coal producers: Mexico, Peru and Chile (see graph 6). Own
production determines increasingly their consumption. Meanwhile their main import trade partner
remains unaltered, being it UK for Chile and Peru, and USA for Mexico. In these countries what we
clearly observe is an import substitution process, which became more noticeable after the war.
120.00%
100.00%
80.00%
Chile
Mexico
Perú
60.00%
40.00%
1929
1930
1928
1927
1926
1925
1924
1923
1922
1921
1919
1920
1917
1918
1915
1916
1914
1913
1911
1912
1910
1909
1907
1908
1905
1906
1903
1904
1901
1902
1899
1900
1897
1898
1895
1896
1894
1893
1892
1891
0.00%
1890
20.00%
Illustration 6: Own production share over coal consumption
Mapping all the results, we can capture the geographical nature of such classifications (see map 1).
Proximity and trade areas seem to explain most of the common features we have identified in that
section. As a conclusion we can not explained the coincidence of imports share trade partners over
time, isolating each country, but we have to integrate them in a more enlarge group of countries. We
cannot understand the whole LA&C as a unique unit, but we also miss information if we break it in
multiple country units.
Illustration 7: Map of coal trade by group of countries.
5 EXPLAINING COAL TRADE: A GRAVITY MODEL
Coal was an essential resource for modernization in the period we focus on. If coal production can
be neglected in the LA&C region, with some specific exceptions, we can assume that trade imports
were the way to achieve modernization. As a consequence, explaining coal trade we also explain
modernization chances. That link weakens after the First World War as long as energy transition to
oil appeared at that time.
We have defined a gravity model as follows:
ln  IMPCOALpc i , j , t =1 2∗ln  EXPTOTG2pci ,t −3∗ln  DST i , j
 4∗ln  ProdCOALi , j ,t 5∗lnQUEUAi ,t i , j ,t
were IMPCOALpci,j,t are the coal imports per capita of each LA&C country from either UK or USA
in a year t. We have only included British and USA coal trade because, as explained before, they
were the main suppliers. We have chosen trade per capita in order to neutralize the scale effect of
considering so different countries in terms of sizes.
EXPTOTG2pci,t is the market effect of the importer country, per capita. We have approached it
through total imports of each LA&C country from UK plus USA. The data used here is the same as
analysed in the first section of the paper (see table 1).
DISTi,j, is the maritime distance between the LA&C country capital city to London or Washington.
We have modified these distances from 1914 on, considering the opening of the Panama Canal.
Distance is supposed to have a negative sign in the equation, more distance being associated to
more transport costs.
QUEUALN is a measure of USA coal trade dependency. It is the share over coal imports from USA
for each year. That variable is the opposite to the British dependency. Low percentages of coal
coming from USA meant higher shares coming from UK, except for the coal producers countries.
In that case, we have introduce an additional explanatory variable: PRDCOAL, which is the coal
production of the country who also imports coal abroad. Results appear in table 7.
Table 7- Estimation results for the coal trade gravity model
Variable
Coeficiente
Desv. típica
Estadístico t
valor p
const
-4,38893
0,497312
-8,8253
< 0,00001
***
EXPTVpc
0,434625
0,0320867
13,5453
< 0,00001
***
PRDCOALp
0,150577
0,0281089
5,3569
< 0,00001
***
QUEUALN
-0,660442
0,0242664
-27,2163
< 0,00001
***
0,128535
0,0461343
2,7861
0,005388
***
DST2
Significance is quite reasonable for all the parameters of the estimation. Total imports explained
coal trade because the more opened of a LA&C country the more chances to modernization it had,
at that time. Distance appears with the expected negative sign, showing the influence of higher
transport costs. Unexpected negative sign is related to the USA share variable. But we can
explained it through the role of the energy transition to oil, mainly supplied by USA. A higher USA
trade dependency was determinant to change from coal to oil.
On the other side, a negative sign for the USA share means a positive sign for the British one. UK
promoted a longer maintenance of coal on the countries it had predominance. The positive sign of
the coal production is also unexpected, due to the import substitution process. But although the
share of coal imports was decreasing over time, total imports in values were increasing. This
explains the positive sign, as long as coal producers had a strong path dependency on coal, instead
of oil.
6 CONCLUSION
We can summarized our results in three main points. First, patterns in total trade differs from that in
coal; basically, the First World War was not a persistent shock in coal trade. British coal exports
manage to recovered most of its LA&C markets in the twenties, meanwhile it dramatically
decreased in the totals. Second, USA coal exports had more geographically diversification than the
British. Its main markets included more countries and they were more scattered; nevertheless
concentration increased during the First Globalization for both suppliers, and decreased after the
First World War. Third, we have find that coal imports were a function of the main supplier and the
participation in international trade. Having USA as a coal supplier, introduced a positive bias to oil,
due to an earlier energy transition11; meanwhile buying coal to UK, favoured coal consumption
persistence. Transports cost, captured by distance, had the usual negative impact. Coal production
was trade promoting trough a coal specialization in the energy structure of consumption.
11 Folchi and Rubio (2006)
BIBLIOGRAPHY
Badia-Miró, M. and Carreras-Marín, A. (2005). Geographical deviations in foreign trade statistics:
a study into European trade with Latin American Countries, 1925. UPF Economics & Business Working
Papers, 884.
Bethell, L. (1986 – 1994). Cambridge history of Latin America. Cambridge, Cambridge University Press,
vols 7 and 11.
Bulmer-Thomas V. (1998). La historia económica de américa latina desde la independencia.
Fondo de Cultura Económica. México DF. México.. .
Carreras, A.,A.A. Hofman,X.Tafunell y C.Yáñez (2003). El desarrollo económico de américa
latina en épocas de globalización. una agenda de investigación. Estudios Estadísticos y
Prospectivos. CEPAL.. .
Folchi, M.; Rubio, M. (2006).
Hofman A. (2000). The economic development of latin america in the twentieth century.
Cheltenham : Elgar. .
Notten, F. (2005).
Thorp R. (1998). Progreso, pobreza y exclusión. una historia económica de américa latina en el
siglo xx.. Washington : Banco Interamericano de Desarrollo[Brusel·les] : Unión Europea, 1998. .
Yañez C, Rubio M & Carreras A. (2006). Economic modernization in the la&c between 1890 1925: a view from the energy consumption..
Official publications
Germany. (Various years). Der Auswärtige Handel Deutschlands. Berlin.
Société des Nations (1928), Mémorandum sur le commerce international et sur les balances des
paiements, 1912-1926, Géneve.
U.S. Department of Commerce. (Various years). The Foreign Commerce and Navigation of the
United States. Washington D.C.
U.S. Department of Commerce. (Various years). Commerce Year Book. Vol. II (Foreign
Countries).
Washington D.C.
United Kingdom: Statistical Office of the Customs and Excise Department. (Various years).
Annual Statement of the Trade of the United Kingdom with Foreign Countries and Britain
possessions. London.