DOI: 10.1111/ehr.13353
ORIG INAL ARTICLE
Trade costs and the integration of BritishWest
Africa in the global economy, c. 1840–1940
Federico Tadei1 Nektarios Aslanidis2 Oscar Martinez2
1Department of Economic History,
Institutions, Policy, and World Economy,
University of Barcelona
2Departament d’ Economia, Universitat
Rovira i Virgili
Correspondence
Federico Tadei
Email: federico.tadei@ub.edu
Funding information
Ministerio de Ciencia e Innovación,
Grant/Award Numbers:
PID2019-105982GB-I00,
PID2022-137382NB-I00,
PID2021-125901NB-I00; MICIU/AEI/
10.13039/501100011033. ERDF/EU;
Ministerio de Economía, Industria y
competitividad, Grant/Award Numbers:
ECO2016-78652-P, ECO2017-88888-P;
Agència de Gestió d’Ajuts Universitaris i
de Recerca, Grant/Award Number:
SGR2021-00729
Abstract
Despite the essential role of trade for African economies, in
the extensive literature on the historical evolution of inter-
national trade costs, Africa is still missing. In this article,
we contribute to filling this gap by (1) providing the first
estimates of British West Africa’s trade costs with Britain
c. 1840–1940 by computing relative price gaps in a repre-
sentative sample of African export and European import
prices, and (2) analysing the main determinants of trade
costs trends, by regressing price gaps on measures of trans-
port costs, market efficiency, and trade barriers. The results
uncover a diverging pattern in African and global trade
costs trends, which was not noticed in the previous liter-
ature. British West Africa experienced a reduction in its
trade costs with Britain c. 1840–70, similar to the one we
observe in other world areas, thanks to improvements in
shipping technology and market efficiency. From the late
1870s, however, as colonial monopsonistic trading compa-
nies consolidated their control of African export markets,
trade costs continued to decline in the rest of the world,
but not in British West Africa. Consequently, from the late
nineteenth century, trade forWest Africa became relatively
more expensive than for other world regions.
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits
use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or
adaptations are made.
© 2024 The Authors. The Economic History Review published by John Wiley & Sons Ltd on behalf of Economic History Society.
474 wileyonlinelibrary.com/journal/ehr Econ. Hist. Rev. 2025;78:474–498.
TRADE COSTS IN BRITISHWEST AFRICA 475
KEYWORDS
Africa, British West Africa, colonization, market integration,
monopsony, price gaps, trade, trade costs
JEL CLASS IF ICAT ION
F1, N7, O43
Africa’s economic development is heavily dependent on the export of primary products.1 Still,
despite the essential role that external trade has historically played forAfrican economies, we have
little systematic knowledge about the role of trade costs in the process of integration of African
export markets into the world economy.2 From the extensive literature on commodity market
integration in the nineteenth and twentieth centuries, we have a good knowledge of international
trade costs between Europe, the Americas and Asia,3 yet Africa is still missing.4 Was Africa differ-
ent compared with the rest of the world? Did Africa suffer from higher international trade costs
than other world regions? And, if this was the case, why?
In this article, we contribute to filling this gap by analysing trade costs of British West Africa
with Britain from the mid-nineteenth century to the eve of the Second World War. We define
trade costs in a broad sense as ‘all costs incurred in getting a good to a final user other than the
marginal cost of producing the good itself’,5 thus including observable components (e.g. shipping,
insurance cost, and duties) and others that aremore difficult tomeasure (e.g. information frictions
and monopoly mark-ups).
1 Deaton, ‘Commodity prices’; Jerven, ‘African post-colonial’; Williamson, ‘Great divergence’; idem, Trade and poverty;
idem, ‘Commodity prices’.
2 Much has been written about the global integration of British West Africa – see the seminal work of MacPhee, British
West Africa and the more recent discussion in Austin, ‘More and more one cog’ and Inikori, ‘British imperialism and
globalization’ – yet we know less about trade costs.
3 Jacks, ‘Intra- and international commodity market’; idem, ‘Commodity market integration’; Jack, Meissner and Novy,
‘Trade costs’; Jacks, O’Rourke and Williamson, ‘Commodity price volatility’.
4 A relatively large literature has focused on inland transport costs, while international trade costs have been somewhat
disregarded. Chaves, Engerman, and Robinson, ‘British West Africa’, analysed wheeled transportation in British Africa.
Bertazzini, ‘Italian colonial roads’, looked at roads in Ethiopia, Somaliland, and Eritrea. Jedwab and Storeygard, ‘Infras-
tructure investment’ constructed a database of railways in Africa between 1862 and 2015. Jedwab, Kerby, and Moradi,
‘History, path dependence’, Jedwab and Moradi, ‘Transportation revolutions’, Okoye, Pongou, and Yokossi, ‘New tech-
nology’, and Herranz-Loncán and Fourie, ‘Public benefit’ focused on railroads in Kenya, Ghana, Nigeria, and the Cape
Colony, respectively. Tadei, ‘The Long-term effects’ and idem, ‘Measuring extractive institutions’ collected estimates of
inland transport costs for French West and Equatorial Africa. Other papers provide information only on specific compo-
nents of international trade costs, such as freights. Limiting our review to Africa, Pascali, ‘Maritime technology’ gathered
data on freight rates between the UK and Algeria, Cabo Verde, Cape of Good Hope, Mauritius, Sierra Leone, and Tunisia
in the second half of the nineteenth century. Federico and Tena-Junguito, ‘World Trade’ collected freight rates between
the UK and Alexandria, Cabo Verde, and Cape Town between 1848 and 1938. Most closely related to our work are Tadei,
‘Measuring extractive institutions’ who estimated international trade costs for French West and Equatorial Africa, and
Tadei, ‘Colonizer identity’, who extended these estimates to British West and East Africa. In both cases, however, the esti-
mates are limited to the first half of the twentieth century, not allowing for the analysis of the evolution of African trade
costs in the long run.
5 Anderson and van Wincoop, ‘Trade costs’, p. 691; Jacks, ‘Intra- and international commodity market’.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
476 TADEI et al.
Our contribution is two-fold. First, we provide novel estimates of trade costs of British West
Africa with Britain from the 1840s to around 1940 by computing relative price gaps in a repre-
sentative sample of British West African export prices and corresponding British import prices.
This lets us assess the development of British West African international trade costs from a
global comparative perspective. We uncover a diverging pattern in the evolution of British West
African trade costs compared with the rest of the world, which was not noted in the previ-
ous literature. International trade costs in British West Africa decreased considerably from the
early 1840s to the 1870s, with a similar decline also being observed in other world regions.
From the 1870s, however, trade costs continued to decline in the rest of the world, but not in
British West Africa, which instead experienced a sharp increase. After the First World War,
trade costs were declining across the globe, and to a smaller extent also in Africa. Yet, from
the 1920s, and particularly in the wake of the Great Depression, British West African trade
costs rose more rapidly than in other world regions. Overall, from the late nineteenth cen-
tury, trade for British West Africa became relatively more expensive than for the rest of the
world.
Second, we investigate the factors behind this divergence. We start by analysing the differ-
ent components of British West African trade costs. Observable factors such as shipping, port
charges, insurance costs, and export duties generally account only for a part of our trade cost
estimates (at most one-half). Instead, other components, which are more difficult to measure
(e.g. inefficiencies, information frictions, monopoly mark-ups, and other trade barriers), seem
to play a more significant role in affecting trade cost dynamics.6 To further explore this result,
we follow Chilosi and Federico7 and regress our price gaps–measure of trade costs on mea-
sures of transport costs, market efficiency, and barriers to trade, such as the presence of colonial
trade monopsonies. We find that the reduction in British West African trade costs between the
1840s and the 1870s can be explained by declining transport costs due to the introduction of
new shipping technologies and by increased market efficiency. The rise in trade costs in the
late 1870s‒80s, and the consequent divergence between Africa and the rest of the world, can
be largely attributed to the activity of monopsonistic trading companies. Their large markups
produced a substantial increase in British West African trade costs, which could not be com-
pensated by the reduction in transport costs and the improvement in efficiency experienced
by West African markets in the same period (introduction of telegraph lines, construction of
railways).
To the best of our knowledge, this paper provides the first systematic estimates of trade costs for
British West Africa and their major determinants from the mid-nineteenth century to the Second
WorldWar, thus allowing us to compare the process of integration in the global economy of Africa
with those experienced by the other world regions.
The rest of the paper is structured as follows. Section I provides the historical and institutional
background regarding West African trade. Section II describes the price data. The paper’s main
results are shown in section III, where we present estimates of British West Africa trade costs
with Britain and draw comparisons between international trade costs for British West Africa and
the rest of the world. In section IV, we explore the determinants of trade costs trends. Section V
provides concluding remarks.
6 Similar results are found in Chilosi and Federico, ‘Early globalizations’ and in Steinwender, ‘Information frictions’.
7 Chilosi and Federico, ‘Early globalizations’.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TRADE COSTS IN BRITISHWEST AFRICA 477
I HISTORICAL BACKGROUND
For West Africa, the nineteenth century marked the development of an agricultural export econ-
omy.8 According to Hopkins, the development of commodity trade, namely palm oil, and in the
second half of the century, groundnuts, represented an important break inWest Africa’s economic
history and indicated the start of modern economic history inWest Africa.9 The commercial tran-
sition from slave to commodity exports led to the entry of new European and small-scale African
producers and traders and the growth of commodity trade made the distribution of incomes
more egalitarian.10 Still, Hopkins notes some key elements of continuity with the slave trade.11
Indeed, during the first half of the nineteenth century, the palm oil trade was organized to a large
degree around the foundations, organization, and practices (port facilities, merchant relations,
and credit instruments) that had characterized the slave trade era.12 Slave labour was used in
the production as well as in the transportation of palm oil.13 Palm oil commerce relied heavily
on shipping (initially on sail but later in the century on steam) and on water transportation by
canoes, making extensive use of broking and trading networks established during the slave trade.
British merchants bought palm oil from the Gold Coast and Nigeria (in particular from the
regions of Bonny, Old Calabar, and Elem Kalabari) and shipped it to major British ports (Liv-
erpool, Bristol, and London) and to a lesser extent to other European ports (Hamburg). By the
1850s, palm oil was one of the two most important West African exports (the other one being
groundnuts, which was traded by French merchants in Western Sudan, British merchants in the
Gambia, and Portuguese ones in Guinea Bissau), while Britain had become the largest importer of
palm oil in the world.14 More generally, estimates show that while Britain and France accounted
for about four-fifths of Europe’s trade with West Africa, Britain’s share was two-thirds to three-
quarters of this total.15 The determinant factor here was the industrialization of Britain and
the soaring demand for raw materials and industrial intermediates (e.g. fuels, oils, and fibres),
several of which were available in British West Africa. Palm oil was used as a lubricant for indus-
trial machinery and railway stocks and became an essential ingredient in the manufacturing of
soap, candles, and tinplate production.16 During this period, the merchant community thrived.
From the middle of the century, together with palm oil, trade in palm kernels also grew, with
8 For palm oil trade, in particular, we refer to Lynn, Palm oil trade, pp. 11‒59, 105‒50, for cocoa, to Austin, Ghana, pp.
46‒56, 236‒249, and 279‒303, while for all West African commodities analysed in this study (cocoa, groundnuts, palm oil,
palm kernels, and rubber), see Hopkins, West Africa, pp. 173‒237. Moreover, Tosh, ‘The cash-crop revolution’ provides
an excellent overview of the expansion of tropical Africa’s agricultural production during 1890‒1930 and Austin, ‘The
Ghanaian cocoa’ a more recent analysis.
9 Hopkins,West Africa.
10 Ibid., pp. 175‒6.
11 Ibid., p. 182. It is worth pointing out that Chamberlin, ‘Bulk exports’ argues that the transition from slave to commodity
exports trade is often referred as the transition to ‘legitimate trade’, but this terminology lacks economic character as it
does not recognize the fact that new extraordinary demand for bulky products entirely changed the trading structures of
West Africa.
12 Lynn, Palm oil trade, pp. 11‒59.
13 Ibid., pp. 51‒2.
14 On groundnuts, see Swindell and Jeng, Groundnut trade, for Gambia, and Hogendorn, Nigerian groundnut exports, for
Northern Nigeria.
15 Hopkins,West Africa, p. 179.
16 Lynn, Palm oil trade, p. 105.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
478 TADEI et al.
Germany (Hamburg) becoming a major importer, suggesting an increasing integration of West
Africa into the world markets.
Until the mid-nineteenth century, there were formidable barriers to entry into the palm oil
trade, which derived from the way it was structured (high start-up capital, high costs for buying
or chartering shipping, expert knowledge and experience of the African market). However, later
in the century, the adoption of steamship propulsion as a new transport technology represented
a major change. In 1852, the first regular shipping line was established by the African Steamship
Company and in 1868 another shipping company, the British & African Steam Navigation Com-
pany, entered the West African market.17 The use of steamships, initially with the high-pressure
compound engine and later on with the triple expansion engine, increased fuel efficiency and
reduced costs remarkably (at the end of the nineteenth century, transporting a ton of palm oil
from Lagos to Liverpool costed around £2 10s).18 The steamer was not only cheaper per ton/km
but, as opposed to sailing, could also guarantee regularity of service, thus higher efficiency. More-
over, the use of the steamship significantly reduced risk in West African trade, opening the palm
oil market to new sources of competition. The steamship services opened West African trade to
independent and smaller-scale traders by providing freight space (meaning that traders no longer
had to own or charter their own ships) and by stopping atmanymore ports, which facilitated trade
along the coast (especially for traders thatwould operate at lowmargins). As a consequence, in the
second half of the nineteenth century, several new trading firms (often of smaller-scale) – British
as well as French and German – but also African merchants (mainly liberated enslaved persons
and their descendants) entered theWest African export trade, generating competition and causing
problems to incumbents.19
To alleviate the competitive pressure, several British companiesmerged into theUnitedAfrican
Company, which started its activities on the West African coast in 1879.20 Such mergers allowed
the new company to obtain a de facto monopsony in the purchase of goods from local African
producers. Despite not being an official monopoly, the company’s ability to negotiate commercial
treaties with local authorities substantially increased with the obtention of the royal charter in
1886, after which it changed its name to Royal Niger Company. In 1889, when the African Associ-
ation entered the market, the monopolistic position of the Royal Niger Company was temporarily
challenged. In 1893, however, the two competing companies undertook market-sharing agree-
ments to establish regions of influence.21 In 1899, the cartel agreementwas extended to include two
other smaller trading companies and further control the West Africa export market. Steamship
companies offering shipping services fromBritain toWestAfrica also exertedmarket power. In the
late 1860s, theAfrican SteamShipCompany and theBritish&African SteamNavigationCompany
17 Rönnbäck and Broberg, Capital and colonialism, p. 177.
18 Lynn, Palm oil trade, p. 109.
19 Hopkins, West Africa, pp. 202‒3; Lynn, Palm oil trade, pp. 137‒8. Dumett, ‘African mercantile entrepreneurship’ notes
the rise of a significant group of smaller-scale African merchants in the Gold Coast as a result of the steamship’s faster
service, larger cargo-carrying capacities, and provision for transporting cargoes in smaller lots in the second half of the
nineteenth century. Nwabughuogu, ‘African middlemen’ provides an in-depth analysis of the decline of wealthy African
middlemen in Eastern Nigeria from the early twentieth century, which marked the beginning of direct trade between the
British merchants and the native producers.
20 Rönnbäck and Broberg, Capital and colonialism, p. 178. Hopkins,West Africa, suggests that the Scramble for Africa of
the 1880s was, in effect, an extension of the firms’ strategies of concentrating ownership. More generally, the imposition
of colonial rule could be seen as one the factor integrating West Africa into global markets (Austin, ‘More and more one
cog’, and Inikori, ‘British imperialism and globalization’), at least up to a point.
21 Rönnbäck and Broberg, Capital and colonialism, pp. 178‒80.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TRADE COSTS IN BRITISHWEST AFRICA 479
agreed to fix freight rates and sailing dates.22 In 1890, ElderDempster&Company obtained control
of all West African British shipping companies, still competing, however, with German compa-
nies.23 Finally, in 1895, it reached an agreement with the main German shipping company, Woer-
mann Linie, to limit competition and establish the West African shipping ‘conference’ cartel.24
The late nineteenth century saw falling palmoil prices and reduced trade value. In the 1890s, the
US tariff policies hit the British tinplatemanufacturers, causing a drop in the demand for palm oil,
and palm kernels became an importantWest African export.25 In Britain, palm kernel oil was used
in the production of the innovative ‘Sunlight Self-Washer’ soap and themanufacture ofmargarine,
while palm kernels became an important food for cattle.26 However, neither the increase in the
value of palm kernel exports nor the increase in palm oil production compensated for the decline
in returns from the palm oil trade27 and trade eventually switched to rubber.28 The last quarter
of the nineteenth century witnessed an impressive but short-lived growth of trade in wild rubber
from trees in the forest area from Guinea to Nigeria. However, around 1906‒7, once the natural
supply of rubber trees approached exhaustion, the rubber trade collapsed. The main problemwas
that rubber was not profitable as a cultivated crop, certainly not compared with cocoa.29
In subsequent years, cocoa cultivation experienced the most spectacular growth of all West
African cash crops. With the increasing commercialization of chocolate production and the
invention of new cocoa presses, it became the most important West African export in the twenti-
eth century. The Gold Coast’s volume of exports increased from a few hundred tons per year in the
1890s tomore than 300 000 tons in 1936/7.30 Overall, from 1911 to the 1970s, the Gold Coast was the
world’s largest producer of cocoa beans.31 Aswith the commerce of palm oil, the cocoa trademade
use of existing African broking and trading networks up until the 1930s. Cocoa cultivation had
a particularly noticeable influence on land occupation. As a result, there were significant orga-
nizational innovations in this period. Cocoa farmers in the Gold Coast and Nigeria, for example,
formed companies or lineage groups to buy land collectively and organize labour recruitment.
Cocoa production earned many farmers prosperity, and the Gold Coast became one of the
wealthiest colonies until independence.32 Technological, telecommunication (e.g. postal ser-
vices), and infrastructural developments (e.g. railways) during the 1900s and 1910s helped spread
cocoa quickly. Railroads decreased inland transportation costs, making cocoa production prof-
itable for exportmarkets. Cocoa output continued to increase until themid-1930s due to increased
capacity installed but peaked since then as cocoa prices dropped after the Great Depression.33
22 Lynn, Palm oil trade, pp. 107‒8.
23 For a history of Elder Dempster in West Africa, see Davies, The trade makers.
24 Hopkins,West Africa, pp. 252‒3.
25 Ibid., pp. 178, 189.
26 Lynn, Palm oil trade, p. 118.
27 Ibid., pp. 120‒121, 123.
28 The volume of British palm oil imports peaked in 1895 with steep increases in early 1850s, early 1870s, and early 1890s.
29 Austin, Ghana, p. 48.
30 Hopkins,West Africa, p. 224.
31 Austin, Ghana, p. 51. On Ghanaian cocoa, see also Hill, Cocoa-farmers.
32 Austin, Ghana, p. 249.
33 Ibid., p. 50. Using the case study of cocoa, Austin, ‘Ghanaian cocoaʼ offers a comprehensive analysis of the ‘cash crop
revolution’ of the late nineteenth and early twentieth centuries that took place in Ghana. In this line of research, Berry,
‘Supply response’ shows that there is little relation between cocoa plantings and Nigeria producers’ terms of trade, imply-
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
480 TADEI et al.
At the same time, the twentieth century saw the consolidation of the monopsonistic power
of exporting trading firms.34 With the First World War, the cartel between the Niger Company
and the African Association collapsed. Still, since the late 1910s, after a sequence of mergers and
acquisitions, concentration increased again, and so did the market power of trading companies.35
In 1919, the AfricanAssociation and two smaller companiesmerged to form theAfrican&Eastern
TradeCorporation. In 1920, theNiger Company (previously RoyalNiger Company)was taken over
by Lever Brothers. The period of monopolistic competition which followed eventually ended in
1929 when the Niger Company merged with the African & Eastern Trade Corporation to create
the United Africa Company (UAC).36 The establishment of the UAC also spelt the end of the
monopoly of the West African shipping conference. During the First World War, Elder Dempster
obtained a full monopoly of shipping as the German firms stopped their operations in Africa,
but after the war, other shipping companies entered the market.37 A new shipping cartel was
established in 1924, but in 1929 the UAC already established its own shipping line.38 Overall, in
the 1930s, the strength of the monopolistic position of the UAC was remarkable, handling alone
about half of all West African trade.39
How could such monopsonies be implemented in West Africa? In British colonies, trade
monopsonies were usually not officially supported by the government, at least until the 1930s,
but were the outcome of market dynamics and political forces. First, West Africa had a long
history of trade and a relatively high level of commercialization, which decreased the operational
expenses for the first trading firms that entered themarket and allowed them to gain largemarket
shares. Second, most of the agricultural production was in the hands of African farmers,40
who, unlike European settlers or plantation companies, had little political power to oppose the
establishment of monopsonies.41 If the entry of smaller companies was limited early on by high
costs, later on, the major trading firms, such as the UAC, forced competitors into confidential
market-sharing agreements. Such arrangements would establish fixed prices that the trading
companies had to offer to African producers and the quantity that each company was allowed
to purchase. If a firm had bought an excess amount, it would have had to provide compensation
to the other ones through penalty fees, which were designed to eliminate any advantage gained
from non-compliance. In such a way, the larger companies were able to transform their dominant
position into a de facto monopsony.42
ing that producers did not greatly rely on crop prices (due to uncertainty about future cocoa prices). Instead, producers
make their planting decisions on the basis of changing opportunity costs and the institutional structure of rural factor
markets. On the beginnings of cocoa cultivation in Nigeria, see Hopkins, ‘Innovation’.
34 For an analysis of the consolidation of shipping and trading monopolies in Nigeria, see Olukoju, ‘Inter-war Nigeria’;
idem, ‘European shipping lines’; and idem,Maritime trade).
35 It is worth noting that the trend towards concentration was not unique to trading and shipping. In the banking sector,
Austin and Chibuike ‘Collusion and competition’ document the collusion between the two major banks in British West
Africa after 1916.
36 Rönnbäck and Broberg, Capital and colonialism, pp. 185‒6.
37 Olukoju ‘Elder Dempster’.
38 Hopkins,West Africa, pp. 252‒3; Rönnbäck and Broberg, Capital and colonialism, p. 187.
39 Hopkins,West Africa, p. 251. On the UAC, also see Fieldhouse,Merchant capital.
40 Hopkins,West Africa, pp. 264, 266, 273.
41 Tadei, ‘Colonizer identity’. In some cases, African farmers attempted forms of collective opposition to European cartels,
themost famous example being the Ghana cocoa ‘holds ups’ (Miles, ‘Rural protest’). Indeed, post-1900, price gaps between
Africa and the UK were on average lower for cocoa in Ghana than for the other colonies/commodities.
42 Bauer,West African trade, pp. 202‒3.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TRADE COSTS IN BRITISHWEST AFRICA 481
In terms of the value chain, local farmers sold their crops to middlemen, who then transported
the goods to marketing places where the trading company acquired them.43 Due to the competi-
tive nature of the internal market (many middlemen could be employed by the same company),
middlemen’s profits were low. Trading companies initially relied on the services of other compa-
nies who shipped the goods to Europe, but later on developed their own shipping lines, with the
difference in price between Africa and Europe being the base for the company’s profits.44
II DATA
Overall, commodity exports played a fundamental role in the development of BritishWest Africa.
Yet, their importance can be fully understood only by analysing the integration of British West
African markets into the global economy. One of the most relevant elements in the process of
market integration is international trade costs. To measure them, we compute relative price gaps,
defined as
𝐺 = (𝑝𝑏 − 𝑝𝑎𝑓)∕𝑝𝑏 (1)
where 𝑝𝑏 is the British import price and 𝑝𝑎𝑓 the African export price. An increase in market
integration implies a lower gap between export and import prices and thus lower trade costs.45
We calculate real price gaps as a share of destination British market prices to obtain a unitless
measure of the trade cost comparable across commodities, countries, and years. The choice of the
commodity price as a deflator for nominal trade costs is in line with the literature.46 Other options
such as gross domestic product (GDP) deflator are not appropriate as they include non-traded
goods and services.47
To calculate price gaps, we use data on cocoa, groundnuts, palm oil, palm kernels, and
rubber exports from the Gambia, the Gold Coast (roughly correspondent to modern-day Ghana),
Nigeria, and Sierra Leone, from the 1840s to 1938. These commodities were the most important
agricultural goods exported from West Africa to Great Britain in the nineteenth and early twen-
tieth centuries. In addition to their representativeness, they were exported from more than one
colony in our sample, so the obtained price gaps are not colony-specific, reducing the possibility
of measurement errors. At the same time, the British colonies accounted for a major part of West
African commerce.48 Similarly, as Britain’s share of world trade was large (an average of about 20
per cent),49 analysing trade flows between West Africa and Britain not only gives us information
43 Hopkins,West Africa, p. 203.
44 For a more detailed discussion of how monopsonies worked and how could be implemented, see Tadei ‘Colonizer
identity’, sect. I.
45 In the regression analysis in sect. IV, we use the log of price gaps as dependent variable. As the log of negative number is
undefined, we remove aminority of negative gaps from the analysis. For consistency, also all other results exclude negative
price gaps.
46 Persson, ‘Transport costs’; ShahMohammed andWilliamson, ‘Freight rates’; Hynes, Jacks, and O’Rourke, ‘Commodity
market disintegration’.
47 Persson, ‘Transport costs’, p. 128.
48 Data from the African Commodity Trade Database (Frankema,Williamson, andWoltjer, ‘West African scramble’) show
that British West Africa accounted for about one-third of all West African exports during the nineteenth century and
around two-thirds in the twentieth century.
49 According to the trade database by Fouquin and Hugot, ‘Two centuries of bilateral trade and gravity data: 1827-2014’,
CEPII working paper no. 2016-14.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
482 TADEI et al.
about the integration of the African colonies into the British imperial economy, but also provides
insight about their integration into the global markets. The focus on West Africa is also justified
because the commercial transition from slave to commodity exports occurred there first and only
later in Central and East Africa.50 Thus, West Africa is the ideal region to analyse to evaluate the
evolution of African trade costs in the export of commodities in the long run. Finally, starting the
analysis in the 1840s coincides with the value of commodity exports fromWest Africa overcoming
for the first time the value of slave exports.51
In particular, we use annual data on cocoa exports fromNigeria (1888‒1938) and the Gold Coast
(1878‒1938); palm oil exports from Sierra Leone (1843‒1938), Nigeria (1863‒1938), and the Gold
Coast (1846‒1938); groundnuts and/or palm kernels from the Gambia (1854‒1938), Sierra Leone
(1850‒1938), Nigeria (1863‒1938), and the Gold Coast (1875‒1938); and rubber from the Gold Coast
(1881‒1938), Nigeria (1886‒1938), and Sierra Leone (1877‒1938).52 African port prices (F.O.B.) are
unit values of the exported item (total export value over total quantity) and are obtained from the
African Commodity Trade Database.53 The original sources are colonial yearly custom statistics,
which reported the total value and quantity of exports by commodity. These recordswere compiled
at local customs offices and subsequently consolidated at the colony level. British prices come
from theWorld Trade Database54 and were originally obtained from Sauerbeck55 for palm oil and
from various issues Annual Statement of the Trade of the United Kingdom for cocoa, rubber, and
nuts and kernels.56 The nature of Sauerbeck prices is unclear but are likely either wholesale or
port/unit values prices, while prices based on the trade statistics are unit values, thus representing
the value of goods at the British port (C.I.F.).57
Figure 1 plots the value of palm oil, cocoa, rubber, palm kernels, and groundnut exports as a
percentage of all agricultural exports from British West Africa. From the late nineteenth century,
these five commodities accounted for more than 90 per cent of the total trade value. Early on,
palm products (oil and kernels) and groundnuts, especially in the 1840s and 1850s, represented
the bulk of British West African exports. Later, at the turn of the century, rubber exports picked
50 Law, The commercial transition; Frankema, Williamson, and Woltjer, ‘West African scramble’.
51 Frankema, Williamson, and Woltjer, ‘West African scramble’, figure 1, p. 234.
52 Palm oil and groundnut West African and UK data began even earlier than in our analysis, see tab. 1.9 in Lynn, Palm oil
trade and fig. 5.1 in Swindell and Jeng, ‘Groundnut trade’. The choice of the sample is due to the unavailability of either
British prices (for groundnuts) or African prices (for palm oil) in the earlier period.
53 Frankema, Williamson, and Woltjer, ‘West African scramble’.
54 Federico and Tena-Junguito, ‘World trade’.
55 Sauerbeck, ‘Prices of commodities’.
56 For palm oil in 1843‒4, we use Lynn, Palm oil trade, p. 29. Federico and Tena-Junguito, ‘World trade’ do not have separate
prices for palm kernels and groundnuts, reporting instead a series for ‘nuts and kernels’. For this reason, for each colony,
we construct a combined African ‘groundnuts and kernels’ series by weighting the groundnuts and palm kernels series by
their exported value. The correlation between the British ‘nuts and kernels’ series and the African ‘groundnuts and palm
kernels’ series is 0.82, the highest correlation among all commodities, suggesting the procedure yields reasonable results.
Moreover, as shown in online app. B, the results are not dependent on different weighting procedures used to construct
the ‘groundnuts and palm kernels’.
57 Federico and Tena-Junguito, ‘World trade, 1800–1938: a new data-set.’ working paper on Economic History, no. 16-01
(2016), Universidad Carlos III de Madrid, Instituto Figuerola, p. 18. Even if reported prices were wholesale prices (thus,
possibly including the cost of transporting goods from the port to the central market, London), this cost must have been
quite small relative to the total trade cost from Africa to the UK (and it would be taken into account by commodity fixed
effects in the regression analysis). On the contrary, Sauerbeck prices did not include duties (see p. 18).
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TRADE COSTS IN BRITISHWEST AFRICA 483
F IGURE 1 Percentage of cocoa, groundnuts, palm oil, palm kernels, and rubber over the total value of
agricultural exports from British West Africa. Notes: Export values in African prices. Source: Created from the
African Commodity Trade Database (Frankema, Williamson, and Woltjer, ‘West African Scramble’). [Colour
figure can be viewed at wileyonlinelibrary.com]
up but only for a short period. As the share of palm kernels declined in the early twentieth century,
cocoa became the most significant commodity export.
Following Chilosi and Federico,58 using price gaps between two markets as a proxy for trade
costs requires the following conditions.
(i) First, there must be actual trade between the two markets, otherwise, there is no arbitrage,
and prices move independently from each other. In our case, this condition is satisfied since
Britain was the main trading partner of British West Africa.59
(ii) Second, the quality of goods must be the same in export and import markets. If this were
not the case, a change in the price gap between West Africa and the UK, due to differential
changes in quality over time, could be mistakenly attributed to variations in trade costs. In
our specific case, however, this is not an issue. For British West African colonies, Britain
was the primary destination market accounting for a major part of their overall exports. At
the same time, most British cocoa, palm oil, and rubber imports came from West Africa.60
58 Chilosi and Federico, ‘Early globalizations’, p. 2.
59 See data from the RiCardo Project by Dedinger and Girard, ‘Exploring trade globalization’.
60 For example, between 1934 and 1938, British cocoa imports from West Africa amounted to around 91% of total British
cocoa imports (see Montgomery and Taylor, Cocoa, table 74, pp. 138‒9). In the case of palm oil and palm kernels, as we
discussed in sect. I, West Africa was the largest producer in the world, with Britain being the main destination market
(Lynn, Palm oil trade, p. 12). For rubber, from 1890 to 1905 the Gold Coast was the largest exporter of rubber in the British
Empire (see Dumett, ‘The rubber trade’). In the case of groundnuts, instead, a smaller proportion of British imports came
fromWestAfrica (in 1937, for example, BritishAfrica andBritish India together accounted for about half of theUK imports,
see United States Department of Commerce, p. 36). Yet the trade costs estimates are unlikely to be affected, as groundnuts
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
484 TADEI et al.
For this reason, variations over time in the average quality of goods in the two markets are
comparable.61
(iii) Third, the traded goods must be representative of trade flows between the two markets. As
shown in figure 1, this requirement is well satisfied, as the five commodities represented a
very large share of total trade.62
Finally, we need to ensure that prices in both Africa and the United Kingdom are reliable. Gen-
erally, British trade statistics are considered quite accurate63 for both import prices in Britain and
export prices from the African colonies. As such, prices in Britain have been widely used in the
literature.64 Similarly, African prices from the British colonies have also been used inworkswhere
their reliability is key.65 Given their accuracy, British prices have been employed as a benchmark
to evaluate the reliability of other countries’ export prices.66 The value of British imports and
exports statistics tend to match the value of corresponding trade flows reported by partner coun-
tries.67 Moreover, British statistics seem sufficiently precise also in the case of developing regions
or colonies68 whose accuracy should be similar to that of African colonies.
To further check the reliability of price data, we look at their correlation with the world price
indexes reported by Jacks69 and with observed African producer prices.70 Such correlations will
not be perfect because of differential changes in trade costs, exchange, and inflation rates over
time, but they should be large enough. The estimates show that prices in the United Kingdom are
highly correlated with world prices, with correlation coefficients ranging from 0.81 in the case of
palm oil to 0.89 in the case of rubber. Similarly, prices at the African ports are highly correlated
with observed African producer prices, with correlation coefficients ranging from 0.79 to 0.97.71
III RESULTS: TRADE COSTS BETWEEN BRITISHWEST AFRICA
AND BRITAIN
Figure 2 shows real price gaps between British West Africa and Britain. To calculate overall price
gaps from our colony/commodity series, we compute the average of the individual series weighted
are, from the late 1860s, only a small portion of total British West Africa trade (see fig. 1). Moreover, as shown in online
app. C, the results are very similar including or excluding the groundnuts and palm kernels series.
61 It is important to acknowledge that this argument is valid only for comparisons of average trade costs over time and that
more caution is required when comparing trade costs between colonies as different price gaps between each colony and
the United Kingdom might be explained by differences in quality across colonies.
62 Exports are non-agricultural goods represented a very small portion of total trade.
63 Lampe, ‘Bilateral trade flows’; Pascali, ‘Maritime technology’.
64 Jacks, O’Rourke, and Williamson, ‘Commodity price volatility’; Federico and Tena-Junguito, ‘World trade’.
65 Frankema, Williamson, and Woltjer, ‘West African scramble’; Tadei, ‘Colonizer identity’.
66 See, for example Tena and Willebald, ‘Export growth’.
67 Federico and Tena-Junguito, ‘Morgenstern revisited’.
68 Carreras-Marín and Badia-Miró, ‘Asignación geográfica’ for Latin America; Federico and Tena-Junguito, ‘Morgenstern
revisited’, pp. 264 and 271 for British India.
69 Jacks, ‘Real commodity prices’.
70 As Jacks, ‘Real commodity prices’ reports indexes of real prices, we calculate nominal indexes using the US CPI from
Williamson, ‘Annual consumer price’.
71 See tab. S1 of the online app. of Tadei, ‘Colonizer identity’.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TRADE COSTS IN BRITISHWEST AFRICA 485
F IGURE 2 Real trade costs of British West Africa with Britain. Notes: Average of colony/commodity price
gaps series, weighted by each colony/commodity export value at African prices. Exponential smoothing. Sources:
Creation from original data from Frankema, Williamson, and Woltjer, ‘West African Scramble’ and Federico and
Tena-Junguito, ‘World trade’. [Colour figure can be viewed at wileyonlinelibrary.com]
by their relative share of total trade value. Averaging has the advantage that potential random
errors in the single price gaps series are likely to cancel each other out. Moreover, weighting
by value can reduce the impact of measurement errors in prices, which are more likely to hap-
pen for lower-quantity commodities due to the unit-value nature of our price data. Finally, as
price gaps tend to be quite volatile, we extract the underlying trade costs trend by smoothing
out our series with exponential smoothing.72 This new series better reflects the change in unob-
served trade cost by reducing the effect of short-term shocks. In online appendix A, we report real
price gaps for each commodity and colony separately. The levels and the trends are similar across
colonies/commodities, suggesting that the dynamics in the evolution of price gaps do not reflect
changes in terms of coverage across goods and colonies throughout the period.73
At the beginning of our period of analysis, British West Africa price gaps were high, hovering
around 40 per cent of the British price. During the second half of the nineteenth century, they
experienced a decline to about 30 per cent until the mid-1870s, partly driven by the introduction
of new shipping technologies such as steamships. From the late 1870s, however, real price gaps
increased, reaching again almost 40 per cent of the British price around the First World War.74
As we will show in section IV, this can be primarily attributed to the beginning of the activities of
monopsonistic companies at the end of the nineteenth century.
72 See Faust and Wright, ‘Forecasting inflation’. In practice, we settled with a smoothing parameter of 0.05 – the same
used by Faust and Wright, ‘Forecasting inflation’ – which is the weight given to current observations as opposed to past
observations (the latter weight is 0.95). In online app. B, we show that the trends are similar if we remove outliers.
73 Colony/commodity dummies can only explain 3% of the variation in log (real price gaps), while year dummies can
explain 21% (adjusted R2).
74 Using data for all colony/commodity trade costs series, we can reject the hypothesis that price gaps in 1880‒1915 are the
same as in 1845‒80.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
486 TADEI et al.
The timing of this decline in trade costs in the initial period coincides with the boom in West
African terms of trade, although the latter peaked a bit later as the terms of trade of French West
African exports (e.g. groundnuts and gum) continued to rise until the mid-1880s.75 The reason
behind this correspondence is clear. Since African prices are set as the difference between global
prices – determined by global demand – and trade costs, a reduction in international trade costs
for British West Africa implied an increase in African export prices. At the same time, the decline
in trade costs reduced the prices in Africa of European imports. As prices of imports declined and
prices of exports increased, terms of trade increased. Thus, the fall in trade costs between 1850 and
the late 1870s could at least partly explain the boom in terms of trade experienced by British West
Africa during the same period.76 Our findings are also consistent with the results by Chilosi and
Federico,77 who link the boom in terms of trade for Asian countries to their increased integration
with the global economy.
How did the trend in British West African trade costs compare with other world regions?
Figure 3 plots our West African price gaps together with price gaps between America and Europe
and Asia and Europe. We compute Asian and American price gaps as the simple average of all
(non-negative) real price gaps in Chilosi and Federico,78 who gathered prices for Asia and (North)
America from a variety of primary and secondary sources. In panel (a), British West Africa trade
costs are compared with global costs, calculated as the simple average of the price gaps series of
America and Asia, and in panel (b), we compare each region separately.
From the mid-nineteenth century to around 1870, trade costs declined for both British West
Africa and the rest of the world. From the 1870s, however, the trends in international and British
West African trade costs diverged. While trade costs in the rest of the world showed a steady
decline until the 1920s (except for during the First World War), in British West Africa they
increased. In the wake of the Great Depression, trade costs rose in all regions, but in British West
Africa, they grew substantially more. These trends are statistically significant and we can reject
the hypothesis of equal British West African and global trade costs after 1880. Overall, it appears
that in periods of declining global trade costs, such as during the First Globalization, BritishWest
Africa’s costs rather increased. At the same time, when international trade costs rose, such as
during the Inter-War period, West Africa’s ones rose even more. As a result, in the period under
analysis, trade for West Africa became relatively more expensive than for other world regions.
IV RESULTS: EXPLAINING TRADE COSTS TRENDS
To explain the trends in BritishWest African price gaps, we start by considering the role of observ-
able factors such as shipping costs, insurance, and export duties.79 We analyse the case of palm oil.
We estimate nominal shipping costs by applying the methodology of Tadei80 to palm oil shipping
75 Frankema, Williamson, and Woltjer, ‘West African scramble’, fig. 4.
76 Underlying the boom in West Africa terms of trade before the 1880s was, to a large extent, also the dramatic decline in
the prices of British manufactured goods as a result of the Industrial Revolution.
77 Chilosi and Federico, ‘Early globalizations’, p. 16.
78 Ibid.
79 British import prices are measured net of import taxes, so such duties are not included in our trade costs estimates.
See Federico and Tena-Junguito, ‘World trade, 1800–1938: a new data-set’, working paper on Economic History, no. 16-01
(2016), Universidad Carlos III de Madrid, Instituto Figuerola, p. 18.
80 Tadei, ‘Measuring extractive institutions’ and idem, ‘Colonizer identity’.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TRADE COSTS IN BRITISHWEST AFRICA 487
F IGURE 3 Comparison with global trade costs. Notes: In panel (a), world trade costs are the simple average
of the average price gaps series of America and Asia. In panel (b), the series are the simple averages of all
available commodity/route series. In computing averages, negative price gaps are excluded. Exponential
smoothing. Sources: British West Africa: see fig. 2; other regions: elaborations from Chilosi and Federico, ‘Early
globalizations’ [Colour figure can be viewed at wileyonlinelibrary.com]
costs per ton/km in base year 1881 from Lynn81 and deflate them by the British palm oil price.82
Our estimates show that shipping costs generally represented a relatively small share of total price
gaps, accounting, at the start of our period, for about 30 per cent. Later on, with the introduction
of the steamship, shipping costs temporarily dropped to around 20 per cent of price gaps to go up
again during World War I and the early 1920s.
81 Lynn, Palm oil trade, p. 109.
82 To estimate nominal costs, the unit cost per ton/km in 1881 is first multiplied by the distance from each colony’s port
to Liverpool. Then, to estimate variation over time, it is multiplied by the freight index from Shah Mohammed and
Williamson, ‘Freight rates’. See Tadei, ‘Measuring extractive institutions’ and Tadei, ‘Colonizer identity’ for details.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
488 TADEI et al.
Insurance also represented a relatively small portion of the trade costs of British West Africa
with Britain. During the first half of the nineteenth century, it accounted for about 15 per cent of
price gaps, since insurance for palm oil was 5–7 per cent of the value of the transported goods83
and trade costs accounted for about 40 per cent of that value (see figure 2). Subsequently, with the
introduction of steam and iron ships, the risk of transporting cargo declined substantially, and so
did the insurance rates. For example, in the case of transatlantic trade, estimates show insurance
rates after the 1920s at 1 per cent of the value of the goods, which would be equivalent to around
5 per cent of price gaps.84
Similarly, the share of trade costs accounted for by export duties was also small. Using informa-
tion from the colonial British Bluebooks, we calculate customs duties for 1903‒38 and find they
represented on average only 5 per cent of price gaps. In the previous period, duties were likely to
account for an even smaller share as taxes on exports were uncommon.85
Overall, observable components of trade costs of British West Africa with Britain such as
shipping, insurance, and export duties accounted for at most 50 per cent of real price gaps.86
Unobservable factors must have played a larger role in affecting trade cost dynamics. This echoes
previous studies examining trade costs for Europe and theUnited States. Steinwender emphasizes
the role of information frictions; Jack, Meissner, and Novy highlight the importance of mone-
tary regimes; and Jacks finds evidence of trade costs being driven by the choice of monetary and
commercial policies more than by changes in freight rates.87
In the case of West Africa, commercial policies were, in fact, an essential determinant of
the structure of export markets. Data for the twentieth century show ample evidence of the
monopsony power of colonial trading companies over West African producers. In French West
Africa, producer prices were substantially lower than they would have been in a competitive
market.88 Similar results can be found for British West Africa, where monopsony profits were
also significant.89 On the basis of these figures, we calculate that monopsony profits amounted
to about 15 per cent of the European price, accounting for about 45 per cent of price gaps.90
To formally investigate the reasons behind the trends in BritishWest African trade costs, we fol-
low the approach by Chilosi and Federico91 and explain variations in real price gaps with changes
in transport costs, barriers to trade, and market efficiency. In particular, we run the following
fixed-effects panel data regression
ln
(
𝐺𝑖,𝑗,𝑡
)
= 𝛼𝑖,𝑗 + 𝛽1𝑇𝑖,𝑗,𝑡 + 𝛽2𝐵𝑡 + 𝛽3𝐸𝑡 + 𝜀𝑖,𝑗,𝑡 (2)
83 Lynn, Palm oil trade, p. 96.
84 Persson ‘Transport costs’, p. 139.
85 Frankema, Williamson, and Woltjer, ‘West African Scramble’, p. 259.
86 Other costs such as storage and spoilage must have been very small for palm oil and cocoa, two highly durable
commodities.
87 Steinwender, ‘Information frictions’; Jack,Meissner, andNovy, ‘Trade costs’; Jacks, ‘Commoditymarket integration’. On
a general note, Anderson and van Wincoop, ‘Trade costs’ suggest that trade costs remain very important even in highly
integrated present-day economies where policy barriers (tariffs) are relatively small.
88 Tadei, ‘Measuring extractive institutions’.
89 Tadei, ‘Colonizer identity’.
90Monopsony profits could also increase if inland trade costs declined (e.g. due to improvements in transportation
technology or new infrastructure) and European trading companies did not raise African producer prices by the same
amount.
91 Chilosi and Federico, ‘Early globalizations’, p. 10.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TRADE COSTS IN BRITISHWEST AFRICA 489
where ln(Gi,j,t) is the logarithm of the real price gap for commodity i, colony j, and year t; αi,j is
a colony/commodity fixed effect; Ti,j,t is the transport costs, measured by the logarithm of a real
freight cost index; Bt is the trade barriers proxied by a set market structure dummies indicating
market concentration among British trading firms; Et is the market efficiency proxied by a linear
yearly time trend; and εi,j,t is the error term. Colony/commodity fixed effects allow us to control
for unobservable time-invariant determinants of trade costs and identify the impact of freights,
barriers to trade, and efficiency within each colony/commodity over time.92
We measure transport costs with a real freight rate index, which varies across colonies, com-
modities, and over time, assuming other transport costs to be proportional to freight.93 To
construct the index, we proceed in the followingway. First, we compute the yearly per cent change
for the nominal freight indexes for all routes from and to the United Kingdom, reported in the
World Trade Historical Database,94 and we average across trade routes to generate a nominal
annual average index. Second, for each commodity, we divide the nominal index by the British
price index (so that both trade and freight costs are in real terms). Third, wemultiply the resulting
real index by each colony’s shipping distance from the UK.95
We proxy barriers to trade with a monopsony dummy, which takes the value of 1 in periods of
strong trading monopsonies and the value of 0 in periods of weak or no trading monopsonies. In
particular, we classify as strong tradingmonopsonies the following periods: 1879‒88 (after the start
of the operations of the BritishUnitedAfrican/Niger Company until the entry of theAfricanAsso-
ciation), 1893‒1913 (after the establishment of the first cartel agreement between the two major
companies until its collapse with the First World War), and after 1929 (when the two main com-
panies merged to create the UAC). Conversely, we classify as weak or no trading monopsonies
the following periods: 1843‒78 (before the United African/Niger Company),96 1889‒92 (after the
AfricanAssociation entered themarket until its agreementwith theNiger Company), and 1914‒28
(the collapse of the cartel agreements during the First World War and the period of competition
which followed).97 Moreover, we control for the effect of shippingmonopolies by adding a dummy
variable for the periods of activity of theWest African shipping conference (1895‒1918 and 1924‒8,
see the historical background in section I).98
We proxy improvements in market efficiency with a yearly time trend, which captures the
introduction of telegraph lines in 1886 in British West Africa99 and the general improvement in
communication and markets over time. The alternative option of using a telegraph dummy is not
feasible as it would be almost perfectly collinear with the post-Berlin Conference period, making
92We also estimated themodel by adding the lag of the endogenous variable using the Arellano–Bondmethod. The lagged
dependent variable is not significant, so we chose the non-dynamic estimation as the main specification.
93 Following Chilosi and Federico, ‘Early globalizations’, p. 7.
94 Federico and Tena-Junguito, ‘World trade’.
95 Shipping distances are calculated using https://sea-distances.org/.
96 The period before the establishment of the United Africa Company in 1879 was characterized by increased competition
in the African trade due to the new shipping lines which allowed smaller traders to enter the market. Indeed, the merging
of several companies into United African Company can be seen as the result of these competitive pressures (Rönnbäck
and Broberg, Capital and colonialism, p. 178).
97 See the historical background in sect. I for details.
98 The differential effect of trade versus shipping monopolies is thus identified by price gap differences in the follow-
ing periods: 1843‒78, 1889‒92, and 1919‒23, no trade nor shipping monopolies; 1879‒88, 1893‒94, and 1929‒38, only trade
monopolies; 1914‒8 and 1924‒8, only shipping monopolies; and 1895‒913, both shipping and trade monopolies.
99 Headrick, The Invisible weapon.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
490 TADEI et al.
it impossible to disentangle the effect of the telegraph and that of colonial rule. We also include
dummies for political and economic events that might have disrupted trade efficiency, such as
formal colonial rule after the Berlin Conference of 1885, the First WorldWar (1914‒8), and periods
of global economic crises (the Long Depression of 1873‒96, the Short Depression of 1920‒1, and
the Great Depression of 1929‒32).100 Moreover, we also control for global price levels by including
a British price index, computed as the average price index of the exported commodities. Finally,
we control for the efficiency of internal markets by including the size of each colony’s railway
network over time.101
Table 1 presents the results.102 The performance of the model is satisfactory. It fits the data rea-
sonably well and the coefficients are stable. Furthermore, classical measurement errors in price
gaps (the dependent variable) would only increase the standard errors, while the same type of
measurement errors in our proxies for transport costs, barriers to trade, and efficiency (indepen-
dent variables) would bias the estimates towards zero, decreasing the likelihood of rejecting the
null hypotheses.
In column (1), we regress the (log) real price gap on the (log) freight rate index, the monopsony
dummy, and the linear time trend. As expected, the freight and the monopsony dummy coeffi-
cients are positive and highly significant. At the same time, market efficiency, measured by the
time trend, is negatively related to trade costs and is statistically significant. In column (2), we
control for global price levels and include dummies for political and economic events that might
have affected trade costs. The coefficient of World War I is positive and significant, consistently
with the increase in trade costs during the war. The depression dummies are significant, indi-
cating the influence of global economic forces on British West African trade costs, but their sign
depends on the specific periods considered. This is not surprising, as economic turmoilmight both
increase the gap between African and UK prices due to the increase in trade costs or reduce it,
given that trading companies would still need to pay African producers aminimumprice for them
to be willing to trade, even in times of lower global prices. This mechanism is also confirmed by
the fact that the price index is positively related to price gaps. The post-Berlin conference period
is associated with higher trade costs, suggesting an additional negative impact of formal colo-
nial rule on top of the effect of trade monopsonies. In column (3), we include a measure of the
efficiency of the internal market, as proxied by the size of each colony’s railways network. Inter-
nal market efficiency is associated with reduced international trade costs.103 In column (4), we
control for monopolies in shipping by adding the West African shipping conference dummy. The
coefficient of the shippingmonopoly dummy is positive, suggesting higher trade costs during that
100 In the UK, the Great Depression ended 1 year earlier than in the United States (Richardson, ‘The economic
significance’).
101We estimate the extension of the railways network as the (log) share of the colony’s ‘cells’, which are less than 10 km
away from a railway line. The data come from Jedwab and Moradi, ‘Transportation revolutions’.
102 As, by construction, both freights and the price index are endogenous, we estimate the regression via IV.We instrument
freights with its lagged value and, following the approach in Chilosi and Federico, ‘Early globalizations’, with the trend
component of a Hodrick–Prescott decomposition of the series of nominal freights divided by the average prices. Similarly,
we instrument the price index with its lagged value and with the trend component of a Hodrick–Prescott decomposition
of the price index. In each specification, among the four possible, we select the set of instruments which maximize the
p-value of the Hansen J overidentification test for exogenous instruments. Nevertheless, as shown in online app. C, the
results are unchanged if we estimate themodel with simpleOLS.We also show that the estimates are similar if wemeasure
the global price level with a United States price index from Jacks, ‘Real commodity prices’, instead of a UK price index.
103 In specification (3), the time trend is not statistically significant. This could be due to that both ‘year’ and ‘railways’ may
capture the same improvement in efficiency over time, making it impossible to precisely estimate their separate effect.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TRADE COSTS IN BRITISHWEST AFRICA 491
TABLE 1 Explaining trade costs.
(1) (2) (3) (4) (5)
Ln (freight) 0.278***
(0.075)
0.254***
(0.071)
0.304***
(0.075)
0.407***
(0.079)
0.287***
(0.087)
Monopsony 0.219***
(0.054)
0.327***
(0.051)
0.189***
(0.053)
0.234***
(0.055)
0.309***
(0.058)
Year −0.008***
(0.001)
−0.008***
(0.002)
−0.002
(0.003)
−0.009***
(0.001)
−0.004
(0.003)
Great depression −0.434***
(0.109)
−0.385***
(0.119)
Short depression −0.497**
(0.209)
−0.488**
(0.207)
Long depression 0.287***
(0.057)
0.244***
(0.067)
Post-Berlin 0.383***
(0.068)
0.309***
(0.072)
WW1 0.285**
(0.135)
0.243
(0.155)
Price level index 0.005***
(0.001)
0.006***
(0.001)
Ln (railways) −0.085***
(0.026)
−0.066**
(0.026)
West African Shipping Confer. 0.256***
(0.048)
0.079
(0.080)
Observations/groups 727/12 727/12 727/12 727/12 727/12
R2 0.0596 0.2012 0.0743 0.0760 0.2070
Under identification (LM test)
p-Value
110.117
0.000
137.778
0.000
111.400
0.000
102.336
0.000
152.680
0.000
Overidentification test
(Hansen J)
p-Value
1.829
0.1762
0.611
0.4345
5.267
0.0217
0.066
0.7978
1.050
0.3055
Note: The dependent variable is Ln (real price gap). All models include colony/commodity fixed effects and are estimated by IV
using a robust variance-covariance matrix.
***p < 0.01; **p < 0.05; *p < 0.1.
period. In column (5), we include all control variables together. The coefficients are very similar
to the previous specifications, but both the time trend and the shipping monopolies dummies are
not significant. For the time trend, this is likely because it aims to proxy for changes in efficiency,
which are already captured by the included dummies for political and economic events and by the
efficiency of the internal markets. In the case of shipping monopolies, their effect is not precisely
estimated and the magnitude of the coefficient is small (shipping monopolies increase trade costs
by about 8 per cent). This is not unexpected, as shipping costs are only a relatively small portion
of total trade costs (20–30 per cent). In columns (2), (3), (4), and (5), the coefficients of the primary
variables of interest (freight, trade monopsonies) do not change much relative to column (1) and
are statistically significant.
Turning to the magnitude of the effects, a 10 per cent decrease in freight is associated with an
almost 3 per cent decrease in trade costs. Increases in market efficiency imply a reduction in trade
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
492 TADEI et al.
costs by less than 1 per cent per year. Monopsonies increased trade costs by around 30 per cent.
These estimates are consistent with Chilosi and Federico who find, albeit in the different context
of Asia, that the monopoly of the East India Company augmented price gaps by about 35 per cent
in the short run and that a 10 per cent reduction in freights reduced trade costs by 3–7 per cent.104
The estimated effect ofmonopsonies is also consistentwith the direct evaluation ofmonopsonistic
profits from Tadei, which suggests that such profits accounted for up to 40 per cent of trade costs
in twentieth-century West Africa.105
To further check the relationship between monopsonies and trade costs of British West Africa
with Britain, we take advantage of the temporal variation in the strength of monopsonies. We
regress log (price gaps) on all significant variables from specification (5) of table 1 and a vec-
tor of monopsonies dummies, one for each of the following periods: 1879‒85 (beginning of the
United African/Niger Company’s activity), 1886‒88 (Royal Charter), 1889‒92 (competition with
the African Association), 1893‒1913 (implementation of cartel agreements), 1914‒28 (cartel dis-
solution and period of monopolistic competition), and 1929‒38 (merger into the UAC).106 The
pre-monopsony period (1843‒78) is the omitted category. This allows us not only to evaluate the
impact of the presence of monopsonies, but also to test hypotheses regarding their relative impor-
tance over time. Figure 4 reports the coefficient of each monopsony period dummy (i.e. their
impact on trade costs relative to the period before the first introduction of monopsonies) and 90
per cent confidence intervals.When comparing dummies across periods, it is important to remem-
ber that their difference might be significant at a 1 per cent or 5 per cent level even if their 90 per
cent confidence intervals overlap (Cumming and Finch ‘Inference’). We will comment on such
differences in the discussion which follows.
The start of the operations of the BritishUnitedAfrican/Niger Company in 1879 should increase
trade costs relative to the previous period and the 1879‒85 dummy is positive and significant. In
1886, the obtention of the Royal Charter allowed the company to consolidate itsmonopsony power
by establishing treaties with the local rulers, and the monopsony coefficient increases (the differ-
ence between the 1879‒85 and the 1886‒8 dummy is significant). After 1889, with the entry of the
African Association, competition augmented and the monopsony coefficient declines (even if the
difference is not statistically significant). After 1893, the two companies established a series of
cartel agreements. The coefficient of the monopsony dummy is not statistically different from
the periods of the royal charter. The dissolution of the cartels with World War I and the period
of competition that followed should imply a reduction in the monopsony dummy, and this is
exactly what we observe: the 1914‒28 dummy is not statistically different from zero, indicating the
absence of monopsonistic profits, and the difference with the previous period is statistically sig-
nificant. Finally, the coefficient increases again and becomes statistically significant in 1929 when
the Niger Company and the African Associationmerged to form the UAC (the difference with the
previous period dummy is statistically significant). Overall, the trend in the monopsony-dummy
coefficient nicely aligns with what we would expect from the historical description of the monop-
sony power trend over time, both in terms of the direction of the changes and their magnitude.
This further emphasizes the importance of monopsonies as a determinant of trade costs trends in
colonial British West Africa.
The results of table 1 tell us how much of the variance in trade costs can be attributed to
transport costs, monopsonies, and efficiency. Yet the most crucial question is to assess their
104 Chilosi and Federico, ‘Early globalizations’, tab. 6, cols. 4 and 6.
105 Tadei ‘Measuring extractive institutions’; idem, ‘Colonizer identity’.
106 See the historical background in sect. I for details.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TRADE COSTS IN BRITISHWEST AFRICA 493
F IGURE 4 Strength of monopsonies over time. Notes: The figure reports the coefficient of the monopsony
dummies (continuous line) and 90% confidence intervals (dotted lines) from the regression of log (trade costs) on
log (freight), year trend, and a vector of monopsonies dummies, one for each period: 1879–85, 1886–8, 1889–92,
1893–1913, 1914–28, and 1929–38 (see text for details). Each coefficient can be interpreted as the increase in trade
costs relative to the period before monopsonies (1843–78). All other variables’ coefficients are consistent with the
estimates of tab. 1. Source: Author creation, see text. [Colour figure can be viewed at wileyonlinelibrary.com]
contribution to the change in trade costs over time. To address this issue, we follow Chilosi
and Federico and perform a decomposition analysis.107 For every pair of decades, we first
compute the change in average price gaps (i.e. the difference between the average price gap in
the 1840s and the average price gap in the 1850s, between the 1850s and the 1860s, and so on).108
Then, we evaluate how much of such change can be attributed to variations in transport costs,
monopsonies (decade-average of the monopsony dummy), and efficiency. For transport costs
and monopsonies, this is computed as the per cent change of the variable times its elasticity
(calculated from specification 1 of table 1) times the initial price gap, divided by the total change
in price gaps. For efficiency, to allow for differential increases or reductions over time instead of
a constant increase as in table 1, we compute the contribution as a residual (i.e. the portion of the
change not accounted for by transport costs or monopsonies).
Figure 5 shows the weighted average net variation in trade cost across colonies/commodities.
For each pair of decades, we calculate the share of the net change in real trade cost due to transport
costs, monopsony, and efficiency for each commodity and colony. Then, we calculate the mean
across all series, weighted by their export values. Each bar represents the average net variation in
British West African trade costs (measured with real price gaps) between decades and howmuch
of this variation can be attributed to changes in transport cost, monopsonies, and efficiency.
107 Chilosi and Federico, ‘Early globalizations’.
108 The change in price gaps is computed as the difference between the predicted value of the dependent variables at time
1 and at time 0, according to specification 1 of tab. 1.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
494 TADEI et al.
F IGURE 5 Explaining trade costs: decomposition analysis, full sample weighted average. Notes: The size of
the three components in each bar represents the share of the net change, not the total variation due to each
component. Source: Author creation, see text. [Colour figure can be viewed at wileyonlinelibrary.com]
The decline in trade costs from the 1840s to the 1870s is mainly explained by transport advance-
ments and increases in efficiency. This is in line with the establishment of a regular steamship
service between Britain and West Africa in 1852.109 Later on, the continued improvement in
transport costs and efficiency could not compensate for the establishment of monopsonies in
the late 1870s, which explains the trade cost increase. In the following decades (the 1890s‒20s),
reductions in transport costs, increases in efficiency, and a decline in monopsony power were
the main drivers of trade cost reductions. The increase in efficiency is consistent with the
telecommunication and transportation advancements that took place in West Africa in this
period. A telegraph cable was laid to link Britain and its West African colonies in 1886110 and
railway networks started expanding in the late nineteenth/early twentieth centuries. Moreover,
the dissolution of trade cartels in the First World War and the period of competition in the
1920s also contributed to reducing trade costs. Finally, from the Great Depression onwards,
increases in real transport costs and especially the strengthening of monopsony power, as
the Niger Company and the African Association merged into the UAC in 1929, played a
significant role.
V CONCLUSIONS
Although there is a plethora of studies on trade costs in Europe, the Americas, and Asia, little
is known about Africa. To fill this gap, in this article we provided novel estimates of trade costs
betweenWest Africa and Britain from the early 1840s to the late 1930s and we analysed the drivers
of their evolution in terms of transport costs, market efficiency, and trade barriers. We show that
109 Lynn, Palm oil trade, p. 105.
110 Headrick, The Invisible weapon.
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TRADE COSTS IN BRITISHWEST AFRICA 495
from the 1840s to around 1870, international trade costs decreased at a similar rate both in British
West Africa and in the rest of the world. Yet, in the subsequent period, the picture changed. Dur-
ing the First Globalization, a period of declining global trade costs, British West African trade
costs rather increased, and, in the Inter-War Years, trade costs increased more in British West
Africa than in the rest of the world. Consequently, from the late nineteenth century, trade for
West Africa became relatively more expensive than for other world regions. Our results suggest
that the decline in trade costs until around 1870 can be explained by reductions in transport costs
and improvements in market efficiency, while the change in trend after 1870 was mostly driven
by the establishment of colonial trade monopsonies, which increased African producers’ cost in
accessing world markets.
ACKNOWLEDGEMENTS
The authors would like to thank Giovanni Federico (Editor), two anonymous reviewers, Christo-
pher Meissner, seminar participants at UC Davis (2019), African Economic History Network
Conference (2019), Athens University of Economics and Business (2019), Bank of Greece
(2019), Conference on Research on Economic Theory and Econometrics (2021), Spanish Eco-
nomic History Congress (2022), and Universidad Carlos III (2023). Special thanks to David
Chilosi for kindly sharing price data for Asia and America. Nektarios Aslanidis acknowledges
the Spanish government’s support through the projects PID2019-105982GB-I00 and PID2022-
137382NB-I00. Oscar Martinez acknowledges the Spanish government’s support through the
research projects ECO2017-88888-P and ECO2016-78652-P and the Generalitat of Catalunya’s sup-
port through the research project SGR2021-00729. Federico Tadei acknowledges funding from
Grant PID2021-125901NB-I00 funded by MICIU/AEI/ 10.13039/501100011033. ERDF/EU.
DATA AVAILAB IL ITY STATEMENT
The data that support the findings of this study are openly available at OpenICPSR: https://doi.
org/10.3886/E200621V1.
ORCID
FedericoTadei https://orcid.org/0000-0003-2509-4827
NektariosAslanidis https://orcid.org/0000-0001-8990-920X
OscarMartinez https://orcid.org/0000-0002-8871-4707
REFERENCES
Anderson, J. and van Wincoop, E., ‘Trade costs’, Journal of Economic Literature, 42 (2004), pp. 691‒751.
Austin, G., Labour, land and capital in Ghana: from slavery to free labour in Asante, 1807‒1956 (Rochester, 2005).
Austin, G., ‘Explaining and evaluating the cash crop revolution in the peasant colonies of tropical Africa,
ca.1890–ca.1930: beyond vent for surplus’, in E. K. Akyeampong, E. H. Bates, N. Nunn and J. A. Robinson, eds.,
Africa’s development in historical perspective (Cambridge and New York, 2014), pp. 295‒320.
Austin, G., ‘Vent for surplus or productivity breakthrough: the Ghanaian cocoa take-off, c. 1890‒1936’, Economic
History Review, 67 (2014), pp. 1035‒1064.
Austin, G., ‘“More and more one cog in the world economic machine”: globalization, development, and African
agency in British West Africa’, in J. Inikori, ed., British imperialism and globalization, c.1840‒1960: essays in
honour of Patrick O’Brien (Woodbridge, 2022), pp. 135‒170.
Austin, G. and Chibuike, U., ‘Collusion and competition in colonial economies: banking in British West Africa,
1916‒1960’, Business History Review, 81 (2007), pp. 1‒26.
Bauer, P. T., West African trade: a study of competition, oligopoly and monopoly in a changing economy
(Cambridge, 1954).
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
496 TADEI et al.
Berry, S., ‘Supply response reconsidered: cocoa in Western Nigeria, 1909‒44’, Journal of Development Studies, 13
(1976), pp. 4‒17.
Bertazzini,M., ‘The long-term impact of Italian colonial roads in theHorn ofAfrica, 1935–2015’, Journal of Economic
Geography, 22(1) (2022), pp. 181‒214.
Canjels, E, Prakash-Canjels G. and Taylor, A. M., ‘Measuring market integration: foreign exchange arbitrage and
the gold standard, 1879‒1913’, Review of Economics and Statistics, 86 (2004), pp. 868‒82.
Carreras-Marín, A. and Badia-Miró, M., ‘La fiabilidad de la asignación geográfica en las estadísticas de comercio
exterior: América Latina y el Caribe (1908–1930)’, Revista de Historia Economica—Journal of Iberian and Latin
American Economic History, 26 (2008), pp. 355‒73.
Chamberlin, C., ‘Bulk exports, trade tiers, regulation and development. An economic approach to the study ofWest
Africa’s “legitimate trade”’, Journal of Economic History, 39 (1979), pp. 419‒38.
Chaves, I. N., Engerman, S. L. and Robinson, J. A., ‘Reinventing the wheel: the economic benefits of wheeled trans-
portation in early colonial British West Africa’, in E. K. Akyeampong, E. H. Bates, N. Nunn and J. A. Robinson,
eds., Africa’s development in historical perspective (Cambridge and New York, 2014), pp. 321‒365.
Chilosi, D. and Federico, G., ‘Early globalizations: the integration of Asia in the world economy, 1800‒1938’,
Explorations in Economic History, 57, no.3 (2015), pp. 1‒18.
Cumming, G. and Finch, S., ‘Inference by eye: confidence intervals and how to read pictures of data’, American
Psychologist, 60, no.3 (2005), pp. 170–80.
Davies, P., The trade makers: elder dempster in West Africa 1852‒1972 (Teaneck, 1973).
Deaton, A., ‘Commodity prices and growth in Africa’, Journal of Economic Perspectives, 13, no. 3 (1999), pp. 23‒40.
Dedinger, B. and Girard, P., ‘Exploring trade globalization in the long run: the RICardo project’,Historical Methods:
A Journal of Quantitative and Interdisciplinary History, 50, no. 1 (2017), pp. 30‒48.
Dumett, R., ‘The rubber trade of the Gold Coast and Asante in the nineteenth century: African innovation and
market responsiveness’, Journal of African History, 12, no. 1 (1971), pp. 79‒101.
Dumett, R., ‘John Sarbah, the Elder, and African mercantile entrepreneurship in the late nineteenth century’,
Journal of African History, 14 (1973), pp. 653‒679.
Faust, J. and Wright, J. H., ‘Forecasting inflation’, in G. Elliott and A. Timmermann, eds., Handbook of economic
forecasting, Vol. 2, Part A (Amsterdam, 2013), pp. 2‒56.
Federico, G., ‘How much do we know about market integration in Europe?’, Economic History Review, 65 (2012),
pp. 470‒497.
Federico, G. and Tena-Junguito, A., ‘On the accuracy of foreign trade statistics (1909‒1935): Morgenstern revisited’,
Explorations in Economic History, 28 (1991), pp. 259‒73.
Federico, G. and Tena-Junguito, A., ‘World trade, 1800‒1938: a new synthesis’, Revista de Historia Económica –
Journal of Iberian and Latin America Economic History, 37 (2019), pp. 9‒41.
Fieldhouse, D., Merchant capital and economic decolonization: The United Africa Company 1929‒1980 (Oxford,
1994).
Fouquin, M. and Hugot, J., ‘Two centuries of bilateral trade and gravity data: 1827–2014’, CEPII working paper no.
2016-14 (2016).
Frankema, E., Williamson, J. and Woltjer, P., ‘An economic rationale for the West African Scramble? The com-
mercial transition and the commodity price boom of 1835‒1885’, Journal of Economic History, 78 (2018), pp.
231‒267.
Gordon, B., ‘Commerce, colonies, and cacao: chocolate in England from introduction to industrialization’, in L. E.
Grivetti and H.-Y. Shapiro, eds., Chocolate: history, culture, and heritage (Hoboken, 2009), chapter 43.
Headrick, D. R., The invisible weapon: telecommunications and international politics, 1851‒1945 (Oxford, 1991).
Herranz-Loncán, A. and Fourie, J., ‘‘For the public benefit?’ Railways in the British Cape Colony’, European Review
of Economic History, 22, no. 1 (2018), pp. 73–100.
Hill, P., The migrant cocoa-farmers of southern Ghana (Cambridge, 1963).
Hogendorn, J. S., Nigerian groundnut exports: origins and early development (Oxford, 1979).
Hopkins, A., ‘Innovation in a colonial context: African origins of the Nigerian cocoa-farming industry, 1880‒1920’,
in D. Clive and A. G. Hopkins, ed., The Imperial Impact: studies in the economic history of Africa and India (1978),
pp. 83‒96, 341‒2.
Hopkins, A., An economic history of West Africa, (New York, 2nd edn., 2019).
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TRADE COSTS IN BRITISHWEST AFRICA 497
Hynes,W., Jacks, D. andO’Rourke, K., ‘Commoditymarket disintegration in the interwar period’,European Review
of Economic History, 16 (2012), pp. 119‒143.
Inikori, J., ‘British imperialism and globalization: British West Africa, 1821–1900’, in J. Inikori, ed., British
imperialism and globalization, c. 1840‒1960: essays in honour of Patrick O’Brien (Woodbridge, 2022), pp. 97‒134.
Jacks, D. S., ‘Intra- and international commodity market integration in the Atlantic economy, 1800–1913’,
Explorations in Economic History, 42, no. 3 (2005), pp. 381‒413.
Jacks, D. S., ‘What drove 19th century commoditymarket integration?’ Explorations in Economic History, 43 (2006),
pp. 383‒412.
Jacks, D., ‘From boom to bust: a typology of real commodity prices in the long run’, Cliometrica, 13 (2019), pp.
201‒20.
Jacks, D. S., Meissner, C. M. and Novy, D., ‘Trade booms, trade busts, and trade costs’, Journal of International
Economics, 83 (2011), pp. 105‒201.
Jacks, D., O’Rourke, K. and Williamson, J., ‘Commodity price volatility and world market integration since 1700’,
Review of Economics and Statistics, 93 (2011), pp. 800‒13.
Jedwab, R. and Moradi, A., ‘The permanent effects of transportation revolutions in poor countries: evidence from
Africa’, Review of Economics and Statistics, 98, no. 2 (2016), pp. 268‒284.
Jedwab, R. and Storeygard, A., ‘Economic and political factors in infrastructure investment: evidence from railroads
and roads in Africa 1960–2015’, Economic History of Developing Regions, 34, no. 3 (2019), pp. 156‒208.
Jedwab, R., Kerby, E. and Moradi, A., ‘History, path dependence, and development: evidence from colonial
railways, settlers, and cities in Kenya’, Economic Journal, 127 no. 603 (2015), pp. 1467‒94.
Jerven, M., ‘The quest for the African dummy: Explaining African post-colonial economic performance revisited’,
Journal of International Development, 23 no. 2 (2011), pp. 288‒307.
Lampe, M., ‘Bilateral trade flows in Europe, 1857–1875: A new dataset’, Research in Economic History, 26 (2008), pp.
81‒155.
Law, R., From slave trade to “legitimate” commerce: The commercial transition in nineteenth-century West Africa
(Cambridge, 1995).
Lynn, M., Commerce and economic change in West Africa: The palm oil trade in the nineteenth century
(Cambridge, 1997).
MacPhee, A., The economic revolution in British West Africa (Milton Park, 1926).
Miles, J., ‘Rural protest in the Gold-Coast: the cocoa hold-ups, 1908‒1938’, in C. Dewey and A. G. Hopkins, eds., The
imperial impact: studies in the economic history of Africa and India (1978), pp. xx–yy.
Montgomery, E. G. and Taylor, A. M.,World trade in cocoa (Washington DC, 1947).
Nwabughuogu, I. A., ‘From wealthy entrepreneurs to petty traders: the decline of African middlemen in eastern
Nigeria, 1900‒1950’, Journal of African History, 23 no. 3 (1982), pp. 365‒79.
Okoye, D., Pongou, R. and Yokossi, T., ‘New technology, better economy? The heterogeneous impact of colonial
railroads in Nigeria’, Journal of Development Economics, 140 (2019), pp. 320‒354.
Olukoju, A., ‘Elder Dempster and the shipping trade of Nigeria during the First World War’, Journal of African
History, 33: 2 (1992), pp. 255‒71.
Olukoju,A., ‘Slamming the “opendoor”: British protectionist fiscal policy in inter-warNigeria’, Itinerario: European
Journal of Overseas History, 23: 2 (1999), pp. 13‒28.
Olukoju, A., ‘“Getting too great a grip”: European shipping lines and BritishWest African lighterage services in the
1930s’, Afrika Zamani, 9 and 10 (2001–2), pp. 19‒40.
Olukoju, A., The ‘Liverpool’ of West Africa: the dynamics and impact of maritime trade in Lagos, 1900‒1950
(Trenton, 2004).
Pascali, L., ‘The wind of change: Maritime technology, trade, and economic development’, American Economic
Review, 107 (2017), pp. 2821‒54.
Persson, G. K., ‘Mind the gap! Transport costs and price convergence in the nineteenth-century Atlantic economy’,
European Review of Economic History, 8 no. 2 (2004), pp. 125‒147.
Richardson, H. W., ‘The economic significance of the depression in Britain’, Journal of Contemporary History, 4 no.
4 (1970), pp. 3–19.
Rönnbäck, K. and Broberg, O., Capital and colonialism: the return on British investments in Africa 1869–1969
(Cham, 2019).
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
498 TADEI et al.
Sauerbeck, A., ‘Prices of commodities and the precious metals 1846‒1885’, Journal of Statistical Society of London,
49 (1886), pp. 581‒648.
Sauerbeck, A., ‘Price of commodities in 1898’, Journal of Statistical Society of London, 62 (1899), pp. 179‒193.
Sauerbeck, A., ‘Price of commodities in 1907’, Journal of Statistical Society of London, 71 (1908), pp. 132‒146.
Sauerbeck, A., ‘Price of commodities in 1913’, Journal of Statistical Society of London, 77 (1914), pp. 556‒570.
Sauerbeck, A., ‘Wholesale price of commodities in 1928’, Journal of Statistical Society of London, 92 (1929), pp.
239‒255.
Sauerbeck, A., ‘Wholesale price of commodities in 1938’, Journal of Statistical Society of London, 102 (1939), pp.
267‒284.
Shah Mohammed, S. andWilliamson J., ‘Freight rates and productivity gains in British tramp shipping 1869‒1950’,
Explorations in Economic History, 41 (2004), pp. 172–203.
Steinwender, C., ‘Real effects of information frictions: when the states and the United Kingdom became United’,
American Economic Review, 108 (2018), pp. 657‒696.
Swindell, K. and Jeng, A.,Migrants, credit and climate: the Gambian groundnut trade 1834‒1934 (Leiden, 2006).
Tadei, F., ‘The long-term effects of extractive institutions: evidence from trade policies in colonial French Africa’,
Economic History of Developing Regions, 33 no. 3 (2018), pp. 183‒208.
Tadei, F., ‘Measuring extractive institutions: colonial trade and price gaps in French Africa’, European Review of
Economic History, 24 no. 1 (2020), pp. 1‒23.
Tadei, F., ‘Colonizer identity and trade in Africa: were the British more favourable to free trade?’, Economic History
Review, 75 (2022), pp. 561‒78.
Tena-Junguito, A. and Willebald, H., ‘On the accuracy of export growth in Argentina 1870‒1913’, Economic History
of Developing Regions, 28 (2013), pp. 28‒68.
Tosh, J., ‘The cash-crop revolution in tropical Africa: an agricultural reappraisal’, African Affairs, 79 (1980), pp.
79‒94.
Williamson, J., ‘Globalization and the great divergence: terms of trade boom, volatility and the poor periphery,
1782‒1913’, European Review of Economic History, 12 no. 3 (2008), pp. 355–391.
Williamson, J., Trade and poverty:when the Third World fell behind (Cambridge, MA, 2011).
Williamson, J., ‘Commodity prices over two centuries: trends, volatility, and impact’, Annual Review of Resource
Economics, 4 (2012), pp. 185–206.
Williamson, S. H., ‘The annual consumer price index for the United States, 1774‒Present,’ MeasuringWorth, http://
www.measuringworth.com/uscpi/
OFF IC IAL PUBL ICAT IONS
United Kingdom. Annual statement of the trade and navigation of the United Kingdomwith foreign countries and
British possessions in the year . . . (later) Annual statement of the trade of the United Kingdom with foreign
countries and British possessions in the year. London: Board of Trade, Statistical Department.
United States Department of Commerce, Foreign Commerce Weekly, vol. V no. 1 (1941).
SUPPORT ING INFORMATION
Additional supporting information can be found online in the Supporting Information section at
the end of this article.
How to cite this article: Tadei, F., Aslanidis, N. and Martinez, O., ‘Trade costs and the
integration of British West Africa in the global economy, c. 1840–1940’, Economic History
Review, 78 (2025), pp. 474–498. https://doi.org/10.1111/ehr.13353
 14680289, 2025, 2, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1111/ehr.13353 by Readcube (Labtiva Inc.), W
iley Online Library on [29/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License