Bo-Dissecting The Sinews of Power - International Trade and The Rise of Britain's Fiscal-Military State, 1689-1823 w30754
Bo-Dissecting The Sinews of Power - International Trade and The Rise of Britain's Fiscal-Military State, 1689-1823 w30754
Bo-Dissecting The Sinews of Power - International Trade and The Rise of Britain's Fiscal-Military State, 1689-1823 w30754
Ernesto Dal Bó
Karolina Hutková
Lukas Leucht
Noam Yuchtman
Niklas Betz, Miguel Ortiz, Davis Kedrosky, and Ekaterina Yudina provided exceptional research
assistance. Helpful and much appreciated suggestions, critiques and encouragement were
provided by James Fenske, Julian Hoppit, David Mitch, Patrick O'Brien, Giorgio Riello, James
Robinson, John Styles, Spike Sweeting, Joachim Voth, Gavin Wright, Nuala Zahedieh, and many
seminar and conference participants. Yuchtman acknowledges financial support from the British
Academy under the Global Professorships program. The views expressed herein are those of the
authors and do not necessarily reflect the views of the National Bureau of Economic Research.
NBER working papers are circulated for discussion and comment purposes. They have not been
peer-reviewed or been subject to the review by the NBER Board of Directors that accompanies
official NBER publications.
© 2022 by Ernesto Dal Bó, Karolina Hutková, Lukas Leucht, and Noam Yuchtman. All rights
reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit
permission provided that full credit, including © notice, is given to the source.
Dissecting the Sinews of Power: International Trade and the Rise of Britain’s Fiscal-Military
State, 1689-1823
Ernesto Dal Bó, Karolina Hutková, Lukas Leucht, and Noam Yuchtman
NBER Working Paper No. 30754
December 2022
JEL No. H2,N43,P1
ABSTRACT
We evaluate the role of taxes on trade in the development of imperial Britain’s fiscal-military
state. Influential work, e.g., Brewer’s (1989) "Sinews of Power," attributed increased fiscal
capacity to the taxation of domestic, rather than traded, goods: excise revenues, coarsely
associated with domestic goods, grew faster than customs revenues. We construct new historical
revenue series disaggregating excise revenues from traded and domestic goods. We find
substantial growth in taxes on traded goods, accounting for over half of indirect taxation around
1800. This challenges the conventional wisdom attributing the development of the British state to
domestic factors: international factors mattered, too.
British expansion from the 17th to the 19th centuries established the largest empire in history and
formed the context for the emergence of modern economic growth. In those centuries, Britain’s
capacity to wage and win wars relied on the expansion of the fiscal capacity of its state (O’Brien,
1988). Where did Britain get the revenue required to pay for its ships and men, and to repay the
debts it incurred in fighting its wars? If revenues were primarily raised from domestic economic
activity, this would support the conventional view among political economists that Britain’s eco-
nomic and political development relied mainly on internal factors (e.g., Britain’s domestic political
institutions and economic dynamism); if revenues were primarily raised from international trade,
this would shift emphasis toward external factors (e.g., trade relying on imperial expansion).
Historical scholarship on the development of the British fiscal-military state — most notably
Brewer’s (1989) Sinews of Power — has emphasized the importance of increased excise tax rev-
enue.1 The excise tax has been treated by many historians as a tax on goods produced (and con-
sumed) domestically; the importance of excise tax revenue thus suggests a central role for revenue
generated from domestic economic activity, as opposed to revenue generated from trade pass-
ing through customs. This account rhymes well with a broader literature that considers domestic
institutions in Britain to be the primary drivers of economic growth (e.g., North and Weingast,
1989; Acemoglu et al., 2005; Acemoglu and Robinson, 2012). Britain’s empire, according to con-
ventional wisdom, was built on its own production, and an efficient internal tax administration
that extracted revenue from it. This conventional wisdom continues to influence the literature on
the political economy of historical development (e.g., Besley and Persson, 2011; Acemoglu and
In this paper, we challenge this conventional wisdom. Although long overlooked, from its
inception the excise was a tax on goods produced and consumed domestically, and also on goods
traded internationally. Indeed, recent scholarship (Hoppit, 2017), while lacking comprehensive
revenue data, suggests that taxes on traded goods like tea, tobacco, and foreign spirits made up a
1 See also Mitchell (1988).
1
significant share of excise tax revenue in the 18th century.2 Evaluating the role of traded goods in
Britain’s increased tax revenues thus requires data more disaggregated than the coarse categories
(e.g., “Customs,” “Excise,” “Stamps,” . . . ) reported in the primary source relied on by the reference
work on Britain’s fiscal development (i.e., Mitchell (1988) and Brewer (1989)).3
We construct such disaggregated data from sources in the National Archives (TNA) collection
“Records of the Board of Customs, Excise and Customs and Excise, and HM Revenue and Cus-
toms” (these are referenced under CUST 145).4 These sources allow us to calculate yearly excise
revenue raised by commodity. Thus, we can decompose the excise revenue according to whether
the good being taxed is produced and consumed domestically (henceforth “domestic”), or instead
produced abroad but consumed domestically, or produced domestically but consumed abroad
(i.e., goods imported or exported; henceforth “traded”). We also construct new disaggregated
customs revenue series that allow us to identify customs on traded goods as well as customs on
domestically-produced coal. To construct the customs series, we relied primarily on the National
Archives collection “HM Treasury – Accounts and Ledgers” as well as the collection “Board of
That domestically-produced coal was taxed at customs is further evidence that the distinction
between customs and excise reflected the management of the tax collection process, not the ori-
gin of the goods being taxed. This is clear when directly consulting the primary sources. These
sources, however, present several challenges (that perhaps contributed to historians’ longstand-
ing reliance on secondary, more aggregate sources). The archival sources often present revenues
at different levels of aggregation, and often have ambiguous labels. For example, CUST 145/22
includes a category labeled “Excise,” which is evidently not all of the traditional excise (i.e., that
reported by Brewer, 1989), because it also lists categories like “soap” and “candles” separately,
which are part of the traditional excise. The “Excise” subcategory in CUST 145/22 is disaggre-
2 The need to disaggregate excise revenues was noted also by O’Brien (1988).
3 The primary source was a compilation produced for Parliament in the second half of the 19th century,“Accounts of
public income and expenditure 1688-1869,” (PP 1868–9, xxxv), referenced by Brewer as “British Parliamentary Papers,
vol. 35 (1868–9).
4 This collection is part of the larger set of documents stored at the British National Archives, “Board of Customs and
Excise and predecessors: Excise Duties, Receipts, Payments and Rates.” We rely primarily on CUST 145/8, CUST 145/12,
CUST 145/18, CUST 145/20, and CUST 145/22.
5 We primarily rely on T 35/55, T 38/357, and CUST 37/50.
2
gated in CUST 145/12. Yet, it is clear that CUST 145/12 alone is insufficient because it leaves out
the categories like soap and candles that are reported in CUST 145/22. Thus, constructing a com-
In some cases, categories may be too broad to allow unambiguous assignment into domestic
or traded categories (e.g., salt and vinegar). We thus construct estimates of disaggregated excise
revenue reflecting conservative assumptions regarding the revenue raised from traded goods. For
example, we compute revenues collected on salt as domestic, though some salt was certainly im-
ported. Nor do we make an effort to decompose the tax revenue on domestically-produced goods
with a traded component to their value added. For example, taxes on domestically processed tex-
tiles are treated as domestic taxation even when the main input (e.g., a less processed textile) was
Even under these conservative assumptions, our data overturn the conventional wisdom re-
garding the importance of domestic taxation versus taxes on trade to Britain’s fiscal development.
The data in Mitchell (1988) and Brewer (1989) suggest that in the early 18th century, taxes on
traded goods represented a minority — around 40% — of total indirect taxes (i.e., taxes on both
traded and domestically produced goods). Over the 18th century and into the early 19th century,
total revenues greatly expanded, and Mitchell (1988) and Brewer (1989) suggest that the tax share
of traded goods fell in this period to around 30% of total indirect taxes. In contrast, our series show
that the tax share of traded goods grew from around 40% of indirect taxation early in the 18th cen-
tury to more than 50% around 1800. During the first quarter of the nineteenth century, traded
goods provided a majority of the revenue from taxed goods. Accounting for revenues collected
from traded goods and then used to subsidize exports and promote other national objectives (i.e.,
revenues spent on “bounties”) slightly increases this share further.7 The increase in revenues from
traded goods from 1689 to the early 19th century accounts for more than half of the overall in-
crease in indirect tax revenues. Tax revenues from overseas trade thus represented a substantial
6 We leave for future work a more complete accounting of the role of trade in the rise of the British state and the
British economy. This would require not only addressing the challenges noted above, but also more precisely estimating
spillovers across sectors and the dynamic consequences of trade for the British economy.
7 The revenues used for bounties did not reach the Exchequer and so were excluded from the revenue figures in
3
component of the fiscal expansion that funded Britain’s imperial dominance.
Our findings have important implications for British and global economic history. Most di-
rectly, we contribute new data on British revenues over time that improve upon the standard
references (Mitchell, 1988; Brewer, 1989). We join Hoppit (2017) in arguing against the traditional
treatment of the excise as taxation of domestic production and provide improved, disaggregated
data on both excise revenue and customs revenue from 1689 to 1823. In so doing, we contribute
fundamental new evidence to the literature analyzing Britain’s fiscal development in the early
modern era (e.g., O’Brien, 2011; Murphy, 2013; Cox, 2016; Dickson, 2017).
In a narrow sense, our disaggregation of the excise allows to connect Britain’s rising fiscal
capacity in the 18th and 19th centuries to specific goods. In particular, we show that the taxation
of products with inelastic demand — so called “drug foods” (Mintz, 1985) — provided a large
share of Britain’s rising tax revenue. Thus, we add to a literature that emphasizes the importance
of trade and colonies to the development of the modern Atlantic economies (e.g., O’Brien, 1982;
Pomeranz, 2000;Acemoglu et al., 2005; Findlay and O’Rourke, 2007; Palma, 2016; Henriques and
In a broader sense, by placing international trade at the center of the fiscal changes experi-
enced in Britain in the early modern era, we contribute to a reassessment of the drivers of British
dominance. Received scholarship explains the rise of modern states by reference to war among
European nations in the 16th to 19th centuries (Tilly et al., 1975; Bonney, 1999; Dincecco, 2011).
Wars mattered because they induced investments in fiscal capacity that could then be used to
fund a growing state that supported the economy (Besley and Persson, 2009). As war became
more costly, it was the states able to raise more revenue that prevailed (Gennaioli and Voth, 2015;
Cantoni et al., 2022). Britain was the winner of that fiscal-military competition. Therefore, iden-
tifying the sources of Britain’s fiscal strength is pivotal to understanding the process of modern
state formation. While wars among European powers produced a need for fiscal capacity for all
contenders, we show that international trade contributed substantial means to building the fiscal
Because the trade that shored up Britain’s fiscal capacity was conducted within the institu-
tional context of the British Empire, our work relates to the literature that emphasizes the role
4
of empire and coercion in the historical development of capitalism (e.g., Williams, 2021; Find-
lay and O’Rourke, 2007;Beckert, 2014; Levy, 2021; Heblich et al., 2022).8 These authors consider
that a strong fiscal-military state helped Britain dominate trade. Findlay and O’Rourke (2007)
also propose that trade fed back into the fiscal-military state through the taxable wealth it cre-
ated. This, they argue, established a mutually reinforcing relationship between economic activity
and the development of the coercive power of the state9 In the case of Britain, the possibility of
such mutually-reinforcing relationship was negated by the conventional wisdom seeing domes-
tic goods as the main contributors to Britain’s fiscal might. Our finding that international trade
provided a substantial share of indirect taxes not only counters the conventional wisdom on fiscal
matters. Our finding also increases the plausibility of explanations in which empire was a driver
In what follows, in Section 2, we describe the role of excise and customs taxes in funding the
British state, particularly in times of war. In Section 3, we discuss the existing historical literature
on the excise tax. In Section 4, we describe the historical data sources we rely on to construct
new, disaggregated excise and customs revenue statistics. In Section 5, we present our newly
Britain’s fiscal capacity, like that of many early-modern European states, was developed in a con-
text of recurrent warfare. In the 17th to 19th centuries, Britain’s wars were increasingly conducted
Brecke (1999) provides comprehensive information on conflicts since 1400. From this data,
we constructed time-series of Britain’s military activity, as well as that of other Atlantic trading
powers: France, the Netherlands, Portugal, and Spain. We also identify which conflicts involve
fighting away from the European continent. From the 18th century on, Britain became the most
belligerent power, and the one most frequently involved in overseas wars. In the 1600–1850 pe-
8 Of course, there were other important contributors to the emergence of the industrial revolution in Britain, from
resource endowments (Allen, 2009); to culture (Mokyr, 2010); to political institutions (North and Weingast, 1989; Ace-
moglu and Robinson, 2012).
9 See also Sánchez de la Sierra, 2020; Acemoglu and Robinson, 2019; Dal Bó et al., 2022; Beraja et al., 2021.
5
riod, Britain fought 273 wars compared to 229 for France, the second most belligerent European
power during the period. Moreover, England is the nation that shifted most aggressively toward
fighting wars overseas. During the period 1600–1700, England fought 39 percent of its wars over-
Britain’s empire was built on winning these wars, that were so frequently overseas, with its
dominant navy. Glete (1993) provides detailed information on the capacity of Britain’s navy and
the navies of its European rivals. We transcribed and harmonized the data on navy strengths in
Glete, 1993, and found that during the period 1600–1800, when Britain developed its fiscal capac-
ity, fought wars, and expanded its empire, its naval strength overtook that of all of its European
rivals. First it overtook the Dutch in the second half of the 17th century, and then the French in the
Such naval superiority was expensive. For example, Findlay and O’Rourke (2007), citing
Baugh (2004), note that “[A] 74-gun ship costing £50,000 to build in 1780 when the largest factory
in England cost only a tenth of that amount.” It is thus unsurprising that each major war Britain
fought was associated with an increase in its stock of government debt (see Online Appendix
Figure A.1). This debt was backed by the promise of government tax revenue, and new taxes
were regularly issued in a manner explicitly linked to the demands of war. For example, in 1689,
Parliament passed “An Act for granting to Their Majesties a Subsidie of Tonnage and Poundage
and other Sums of Money payable upon Merchandizes Exported and Imported” (2 W&M, sess. 1,
cap. 4). Parliament passed this bill “for the better enabling your Majestyes to prosecute the present
Warr against the French King and for the reduceing of Ireland.”
motives
The conventional wisdom on the excise tax is built on three pillars, all of which are well-
summarized by Brewer (1989). First, its domestic scope: Brewer (1989, p. 56) writes that, “The
excise was an indirect commodity tax on domestically produced goods, levied either at their point
of production or distribution.” Second, compared to the customs tax, its great and increasing rel-
6
ative importance: Brewer (1989, p. 80) presents data showing approximately equal levels of excise
and customs revenues collected around 1700, and excise revenue levels that are more than dou-
ble customs revenues in the late 1700s. Third, its contribution to state development: Brewer (1989,
p. 56) writes that, “Excises became the largest category of taxes, excisemen the biggest body of of-
ficials, and the Excise Office a byword for administrative efficiency. . . . [T]he English Excise more
closely approximated to Max Weber’s idea of bureaucracy than any other government agency in
eighteenth-century Europe.”10 There is some truth in this conventional wisdom: excise revenues
did grow throughout the 18th century; the excise administration was an early example of an effi-
However, the conventional wisdom is incorrect in viewing the excise tax or the growth in
excise revenues as entirely driven by domestic forces. This is evident, qualitatively, in the first
excise bill passed by Parliament in 1643, which imposed a tax on, “[A]ll and every the Merchants
and Importers of the said Forraign Commodities in the said Schedule mentioned.”11 In a history
of the excise, the Boards of Customs and Excise describe how “At the Accession of James II [in
1685], the Temporary Excises were renewed for his life, and increased by additional duties on
Wine, Vinegar, Tobacco, and Sugar.” Needless to say, Britain did not produce tobacco or sugar
domestically in 1685.12
To move beyond this initial assessment, we more systematically examine the military motives
behind tax bills, as well as the importance of taxes on trade during the later Stuart reigns (those of
William & Mary and of Queen Anne), when the excise and customs regimes of the 18th century
were established.13 We read and classify every tax bill in 1689–1714 according to its mention of
war and/or trade. Roughly 40 percent of bills across both rulers were “Public,” and could involve
10 Brewer is not alone in taking any of these positions. Beckett and Turner (1990) and Ashworth and Ashworth
(2003) join Brewer in treating excise as a tax on domestic production. The data series on customs and excise taxation
in Mitchell (1988) are constructed from precisely the same coarse historical source as Brewer: the “Accounts of public
income and expenditure 1688-1869.” The efficiency and importance of the excise administration are emphasized by
Coffman (2013).
11 Emphasis added by the authors. The bill is TNA/CUST 145/15, “An Ordnance of the Lords and Commons,
In Parliament, for the speedy Raising and Levying of Monies by Way of Charge and New Impost, upon the several
Commodities in a Schedule annexed,” September 11, 1643.
12 The history of the excise quoted is CUST 155/7, “Some Account of the Excise Duties,” 1829.
13 Studying the Hanoverian monarchs of the 18th century is less revealing. Because the systems of excise and customs
were already established, there were fewer tax bills. In addition, because the purpose of taxation — to pay for war —
became self-evident, it also became implicit, rather than explicit, in tax bills.
7
Table 1: Classification of Tax Bills in 1689-1714: Financing Wars and Taxing Trade
Reign: Bills Private Public Share of Public War War & Trade
1689-1702: William & Mary 807 0.58 0.42 0.22 0.81 0.36
1702-1714: Queen Anne 943 0.64 0.36 0.20 0.77 0.50
Total: 1750 0.61 0.39 0.21 0.79 0.43
Note: This table reports the classification of bills passed during the reigns of William & Mary and Queen Anne. The
coding is based on the authors’ reading of the bills. Column 1 reports the count of all bills in each reign. Columns 2
and 3 decompose the count into the share of private and public bills respectively. Column 4 reports the share of public
bills that are tax bills, column 5 reports the share of tax bills that mention military aims, and column 6 reports the
share of tax bills that mention military aims and also include traded goods. Bills from the reign of William III are
included in row 1 with William & Mary. See section 3 for a discusion of the coding.
matters of taxation.14 We find that around 80% of public tax bills mention military, colonial, or
defense (i.e., “war-related”) objectives in their text; during the rule of William and Mary, 36% of
tax bills mentioned both war and taxes on traded goods, and this simultaneous mention rises to
One may still worry that even if many excise bills mentioned trade, traded goods could still
have been marginal to the excise. Hoppit (2017) has collected evidence suggesting not only that
excise taxes were collected on traded goods as Britain’s fiscal capacity expanded, but also that the
role of traded goods was substantial and growing. Hoppit (2017, p. 293) presents data showing
that in 1741, of the excise revenue collected in London (one-third of all British excise), imported tea
and liquors accounted for nearly 40%. In 1796, imported tea, spirits, wine, and tobacco and snuff
accounted for nearly two-thirds of London’s excise revenue. These data points, as well as our
analysis of excise legislation, suggest the need to re-examine the historical evidence that sustains
the conventional wisdom. As noted above, such a re-examination requires disaggregated data
from archival sources that have not yet been systematically used.
14 “Private” bills, in turn, affected some particular interest more circumscribed than the general public. Examples
are bills affecting communal rights of passage, or roads.
8
4 Constructing new series of excise and customs revenues
To construct our disaggregated excise tax dataset, we begin with CUST 145/22. This source is
sufficient to construct, by individual good, yearly revenue series from 1788 onward, allowing us to
classify revenues as originating in trade or from domestic production. Prior to 1788, CUST 145/22
is not fully disaggregated. It presents good-level excise and inland revenues for many goods that
fall under the traditional heading of “excise.” These include glass, soap, paper, tea, and chocolate,
among others (see Online Appendix Figure A.2 for an image of CUST 145/22). Unfortunately,
CUST 145/22 also includes a category labeled “Excise” which requires further disaggregation. It is
a subcategory of what is traditionally regarded as excise, and it aggregates revenues from different
types of alcohol.
To disaggregate the “Excise” category from CUST 145/22, we turn to CUST 145/8 and
CUST 145/12 for the years up to 1787.15 These sources include revenues information on various
categories of alcohol, which add up to the “Excise” category from CUST 145/22. However they
do not include the other disaggregated revenues that are reported in CUST 145/22 (glass, soap,
paper, tea, etc.), meaning that we need to combine information from CUST 145/22, CUST 145/8,
The next step is to convert the revenues information on various types of alcohol from
CUST 145/8 and CUST 145/12 into revenue data by good at a disaggregated enough level to allow
assignment to traded or domestic categories. In some cases, e.g., “British Spirits”, this can be done
directly from the source. However, most of the revenue reported in CUST 145/8 and CUST 145/12
is organized not according to good, but according to the acts under which taxes were collected
(e.g., “IX Continued quarto Annae”) or allocated (e.g., “ Hereditary and Temporary Excise”).
To convert act-level revenues into revenues by good, we first identify which goods are taxed
under a given act. Then, we rely on data on taxed quantities by good and year, as well as tax rates by
15 CUST 145/8 and CUST 145/12 include the same information; we rely on both sources to overcome the challenge of
illegible documents (see Online Appendix Figure A.3 for an image of CUST 145/12).
9
good and year to calculate the revenues by good and year that fall under a particular act.16 When
rates are ambiguous (for example, imported brandy might be taxed under the French brandy rate
or as generic foreign brandy) we assign the lower rate to calculate the revenue from traded goods
conservatively.
A final obstacle in identifying revenue from traded versus domestic goods is the temporary
revenue category, “P Cent” (a percentage tax temporarily levied on selected commodities). This
category of excise appeared in the CUST 145/22 series between 1779 and 1787 and included both
traded and domestic goods. To disaggregate the category, we use CUST 145/20, which shows the
yearly contributions of each of the goods charged “P Cent” duties (e.g., tea and foreign spirits,
among others).
The vast majority of customs revenues were collected from traded goods, as one would expect.
The primary domestic good that was taxed under customs was coal. In addition, other domesti-
cally produced goods were taxed under customs as “carried coastways goods.” We assign these
revenues to domestic production, along with taxes levied on coal.17 In addition, from 1786–1806,
a tax on windows (under the Commutation Act of 1784) was included in the customs revenues,
Disaggregated customs revenues data come from multiple historical sources. To disaggregate
total customs revenues into revenues from traded goods and revenues from domestically pro-
duced goods, we mainly rely on T 35/55, CUST 37/1, T 38/357, and CUST 37/50. These series end
in 1806. We supplement with detailed tables on revenues by good or by act from the Parliamentary
16 This is not always trivial: for example, rum is in some years implicitly taxed at the same rate as imported brandy,
and sometimes taxed as a distinct commodity. For quantities of taxed goods we relied on CUST 145/20. For rates,
we relied on CUST 145/3, CUST 145/4, CUST 145/11, CUST 145/12, CUST 145/18, and CUST 145/20, as well as printed
primary sources and Parliamentary bills (e.g., Crouch, 1731, Baldwin, 1770, and 6 Geo. II, cap. 17). It is important to
note that this “bottom up” approach may miss some revenues (e.g., temporary excises). Such measurement error is
likely small, however: in Figure 1, below, we show that our aggregate Excise and Custom revenues are extremely close
to those in (Mitchell, 1988) and Brewer (1989).
17 While the vast majority of customs revenues collected from coal are on domestic consumption, a small share of
customs revenues are collected from the international export of coal. We count the latter revenues as revenues on traded
goods.
18 The tax on windows is included under customs because it was enacted alongside a reduction in the tax rate on tea
10
Papers for 1807-1818.
We first identified goods that were undoubtedly traded internationally and taxed under the excise.
Some of these goods, like foreign spirits, are labeled as such. Others are not labeled as foreign, but
were certainly produced outside Britain, like tea, coffee and cocoa nuts, tobacco, and pepper. We
also treat wine as traded - in contrast with “low wine”, which was recorded separately, and which
This makes for an extremely conservative calculation of revenues from traded goods — a lower
bound. When it is possible that a positive share of a good may have been domestic, we assign it
to the domestic category. For example, we do not include in our estimated revenues from trade
those revenues collected on hides or salt, though some hides and salt were certainly imported. In
addition, taxes on domestically processed textiles are treated as domestic taxation even when the
raw input (e.g., a less processed textile, or raw materials like silk or cotton) was produced abroad
We follow the approach of Mitchell (1988) and Brewer (1989) in that we report revenues net of:
(i) the costs of running the respective tax administrations (i.e. “management costs”); (ii) refunds
on import duties paid to re-exporters (called “drawbacks”); and, (iii) revenues used directly to
subsidize domestic producers’ exports or to pay for other national objectives (called “bounties”)
that never reached the Exchequer. To be precise, we collect information on the “Payments into
While netting out the cost of the tax administration matches standard practice, and while
refunds paid to re-exporters do not directly contribute to the strategic policy aims of the state,
bounty revenues were used to support the state’s policy aims. Bounty revenues collected by cus-
toms officers from traded goods and used to pay domestic corn producers or to support the civil
government of Scotland (to give two prominent examples) may not have reached the Exchequer,
but they supported the state’s strategic objectives, nonetheless. Information on these bounty pay-
19 We also follow Mitchell (1988) and Brewer (1989) in reporting revenues in millions of pounds in nominal terms.
It is worth noting that the revenue increases over the time period we cover were not driven by higher price levels.
Inflation over the period studied was low — below one percent per year (Thomas and Dimsdale, 2017).
11
ments were not available in the primary sources consulted by Mitchell (1988) and Brewer (1989),
but we are able to identify this category of revenue, collected from both traded and domestic
goods. We thus construct separate series of revenues from traded and domestic goods that in-
clude the bounty revenues in addition to the “Payments into Exchequer.” In order to keep our
exposition as close as possible to that in the received literature, in the main text we focus on rev-
enue series corresponding to “Payments into Exchequer,” and include bounties in series reported
5 Empirical patterns
We begin by comparing our total excise revenue and customs revenue series for the years 1689–
1823 with those in Mitchell (1988), the standard reference (which is based on the same historical
source as Brewer, 1989). In Figure 1, Panel A, one can see that our construction of total excise
revenue closely matches the aggregate excise data reported previously by Mitchell (1988)).20 In-
cluding the bounties in our excise data series has minimal effect (see Online Appendix Figure
A.4). In Figure 1, Panel B, one can see that our construction of total customs revenue again closely
matches aggregate data that have previously been collected (Mitchell, 1988). Adding the bounties
to our customs data series has a more noticeable effect, but again the broad patterns of total rev-
enue match those in Mitchell (1988). It is worth emphasizing that our data come from a different
set of far more disaggregated historical sources; it is reassuring that these data yield aggregate
12
Figure 1: Comparison of Aggregate Excise and Customs Revenues with Mitchell (1988)
Panel A: Excise Revenues
British excise revenue in £1,000,000
26
24
22
20
18
16
14
12
10
8
Mitchell (1988)
6
4
Our Data
2
0
1689 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1823
Year
11
10
3 Mitchell (1988)
2 Our Data
1
0
1689 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1818
Year
Note: This figure compares the aggregate patterns of excise and customs revenues of the British government (in
£1,000,000) as calculated by the authors (black, solid lines) with those reported in Mitchell (1988) (as grey, dashed
lines). Panel A reports the excise revenues for 1689-1823, and Panel B reports the customs revenues for 1689-1818. Gaps
in the lines indicate years with missing data. See Section 4 for a description of the data and methodology. See Online
Appendix Figure A.4 for a version of this figure which includes revenues spent to finance bounties.
13
5.2 New vs. old series: comparison of disaggregated data
We next decompose aggregate revenues into revenues collected from traded goods and revenues
collected from domestic goods. Let RevenueTradet denote the tax revenue raised on internationally
traded goods, RevenueDomestict denote the tax revenue raised on domestic goods, Customst the
tax revenue collected by customs, and Exciset the tax revenue collected as excise, all during year t.
The approach in Brewer (1989) (using the same data as Mitchell, 1988), is to assume the following
RevenueDomestict = Exciset .
Then, following Brewer (1989), the trade-related share of total indirect tax revenue (i.e., the sum
Customst
ShareTradet = Customst + Exciset .
But as we have argued, it is incorrect to equate taxes on traded goods to taxes generated by
customs, and taxes on domestic goods to the excise. It is necessary to define CustomsCoalt to
denote tax revenue collected by customs on coal (a domestic product), ExciseDomestict to denote
excise taxes raised on domestic goods, and ExciseTradet to denote excise taxes on traded goods.
Then, using our disaggregated excise and customs data we can calculate:
RevenueTradet
ShareTradet = RevenueTradet + RevenueDomestict .
Importantly, we can construct these time series restricting revenues to those paid to the Exchequer
(as in Mitchell, 1988, and Brewer, 1989) or including also the revenues used to pay bounties.
In Figure 2, Panel A, we first show the levels of revenues from customs and excise as pre-
sented in Mitchell (1988) and Brewer (1989). Revenues are stacked on top of each other to sum to
14
total revenue from indirect taxes. One can see two patterns. First, that excise revenue is substan-
tially larger than customs revenue throughout the time period. Second, in the Mitchell (1988) and
Brewer (1989) treatment, taxes on “domestic production” — to be precise, excise revenues — ac-
count for the bulk of enormous increase in revenues in the late 18th and early 19th centuries.
In Figure 2, Panel B, we show the levels of revenues from traded and domestic goods as we
calculate them.22 A very different pattern of revenue growth appears. Revenue from traded goods
increases, rather than declines in importance as the British state developed over the 18th century.
At the height of the Napoleonic wars in the early 19th century, traded goods provide more revenue
than domestic goods. If we take the entire period under study, increasing revenues from traded
goods in 1689-1818 account for 54% of the overall increase of excise and customs revenues. The
growth of the British fiscal military state was not financed on the taxation of domestic goods
alone. Rather, tax revenues from international trade represented a substantial component of the
15
Figure 2: Decomposition of Excise and Customs Revenues
Panel A: Revenues from Customs and Excise (as in Mitchell, 1988, and Brewer, 1989)
British government revenue in £1,000,000
35
30
25
20
15
10
Customs
5 Excise
0
1689 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1818
Year
Panel B: Revenues from Traded Goods and Domestic Goods (as calculated by the authors)
30
25
20
15
10
Traded
5
Domestic
0
1689 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1818
Year
Note: This figure decomposes the excise and customs revenue of the British government (in £1,000,000). Panel A de-
composes the revenues as presented in Mitchell (1988) and Brewer (1989) into revenues from customs and excise. Panel
B plots the levels of revenues from traded goods and domestic goods as calculated by the authors. Years with missing
customs data are linearly interpolated. See sections 4 and 5.2 for a description of the data and methodology. See Online
Appendix Figure A.5 for a version of Panel B which includes revenues spent to finance bounties.
16
In Figure 3, we plot the share of British customs and excise revenue from traded goods cal-
culated using the approach and data in Mitchell (1988) and Brewer (1989) (i.e., treating customs
revenue as coming from traded goods and excise as coming from domestic production), as well
as the share of revenue from traded goods calculated using our disaggregated customs and ex-
cise data. One can see in the figure that the traditional narrative of a modest and declining role
for taxes on traded goods as Britain expanded its fiscal capacity is overturned when examining
disaggregated data on the excise. Indeed, our data show that as revenues expanded enormously
over the second half of the 18th century, the share of revenues from traded goods actually increased
and was over 50% of total excise and customs revenues in the late 18th and early 19th centuries.
The share of revenues from traded goods reaches a peak of over 60% in 1800, when additionally
.65
.6
.55
.5
.45
Our Data
.4
.35
.3
Mitchell (1988)
.25
.2
1689 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1818
Year
Note: This figure compares the share of British customs and excise revenue from traded goods as computed by the
authors with the share as reported in Brewer (1989) and using the data in Mitchell (1988). The grey, dashed line plots
the share following Brewer (1989) in treating customs revenue as coming from traded goods and excise as coming
from domestic production. The black, solid line plots the share of revenue from traded goods following the authors’
calculations and using disaggregated customs and excise data. Gaps in the lines indicate years with missing data. See
sections 4 and 5.2 for a description of the data and methodology. See Online Appendix Figure A.6 for a version of this
figure which includes revenues spent to finance bounties.
17
5.3 What traded goods contributed to fiscal revenues?
An obvious threat to the success of the excise tax would have been charging rates that discouraged
imports. A standard result in public finance due to Ramsey (1927) is that optimal taxes should bear
an inverse relationship to the elasticity of demand. To gain insight into what made a high fiscal
revenue possible we examine the nature of the traded goods that were taxed.
In Panel A of Figure 4, we further decompose excise revenues from trade by good. Tea and
foreign spirits were the most important components throughout the 18th and early 19th centuries,
with wine and tobacco playing an increasingly important role in the early 19th century. The other
goods are cocoa, chocolate, coffee and pepper. The entirety of traded excise goods are consump-
tion items that create habituation and have been noted to have relatively inelastic demands — ac-
tual drugs like alcohol and tobacco, as well as what Mintz (1985) called “drug foods.”23 Panel
B of Figure 4 decomposes the customs revenues from traded goods.24 The set of goods includes
imports that were taxed under both excise and customs (foreign spirits, tea, tobacco, etc.). The
most important contributor to customs revenue was sugar, which was not taxed under excise,
and which accounted for a third of customs revenues. A majority of customs revenues also came
from habituation goods with highly inelastic demand. These findings help us understand why
the combination of taxation and trade activity provided a high volume of fiscal revenue.
23 Pomeranz (2000) leverages Mintz’s characterization to argue that the trade on these goods gave the British econ-
omy an additional boost by expanding labor supply: to afford these goods, individuals altered their labor-leisure
choices toward longer working hours (see also De Vries, 1994).
24 Before 1787 our customs data is disaggregated by tax act, which does not allow for a simple decomposition by
goods.
18
Figure 4: Decomposition of Revenues from Traded Goods
Panel A: Excise Revenues, 1720-1823
British excise revenue in £1,000,000
10
7
Pepper
6 Tobacco and Snuff
Wine
5
Coffee, Cocoa Nuts, Chocolate
4 Tea
Foreign Spirits
3
0
1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1823
Year
12
11
Other Habituation Goods
10 Tobacco
Wine
9
Tea, Coffee and Cocoa Nuts
8 Foreign Spirits
Sugar
7
Other Imports
6 Exports
5
0
1787 1790 1795 1800 1805 1810 1815 1818
Year
Note: This figure decomposes British government revenues from traded goods (in £1,000,000). Panel A plots excise
revenues by traded good for 1720-1823. Panel B plots customs revenues by traded good for 1787-1818. Excise rev-
enues from tea and wine include revenues from tea and wine licences. Excise revenues from imported beer have
been included with the revenues from foreign spirits, since revenues from imported beer are too small to be visible
independently. Customs revenues from other habituation goods come from opium, licorice, pepper, spices etc. The
”Other Imports” category of Panel B includes all the customs revenues from imports other than the habituation goods
enumerated in this figure. See section 4 for a description of the data and methodology.
19
6 Conclusion
We provide new data on British excise and customs revenues over time that improve upon stan-
dard references. The data make clear that: (i) excise taxation was not only a tax on domestically
produced goods, but was also a tax on traded goods; (ii) taxes on traded goods were a large share
of indirect taxation, ranging from 40 to 55%; and (iii) taxes on traded goods were a growing share
of total revenues from the early 18th century to the early 19th century, as Britain’s fiscal-military
state developed. These results are evident even under the conservative assumptions guiding the
The patterns revealed by the data we collect should change the narrative regarding Britain’s
fiscal development. This was not simply a process that relied on domestic production, bureau-
cracy, and taxation. Rather, it relied to a significant extent on the expansion of trade that was built
on mercantilism, war, and empire. The more general implication is that the coercive power of the
state can be both an input to, and an outcome of, taxable economic activity.
20
References
Acemoglu, Daron and James A. Robinson, Why Nations Fail: the Origins of Power, Prosperity, and
Poverty, New York: Crown Business, August 2012.
and , The Narrow Corridor: States, Societies, and the Fate of Liberty, New York: Penguin Press,
2019.
, Simon Johnson, and James A. Robinson, “The Rise of Europe: Atlantic Trade, Institutional
Change, and Economic Growth,” American Economic Review, June 2005, 95 (3), 547–579.
Allen, Robert C, The British industrial revolution in global perspective, Cambridge University Press,
2009.
Angelucci, Charles, Simone Meraglia, and Nico Voigtländer, “How merchant towns shaped
parliaments: From the norman conquest of England to the great reform act,” Technical Report,
National Bureau of Economic Research 2022.
Ashworth, William J and William Ashworth, Customs and excise: trade, production, and consumption
in England, 1640-1845, Oxford University Press on Demand, 2003.
Baugh, Daniel A, “Naval power: what gave the British navy superiority?,” in Leandro Pra-
dos de la Escosura, ed., Exceptionalism and Industrialisation, Cambridge University Press, 2004,
pp. 235–258.
Beckett, John V and Michael Turner, “Taxation and economic growth in eighteenth-century Eng-
land 1,” The Economic History Review, 1990, 43 (3), 377–403.
Beraja, Martin, Andrew Kao, David Y Yang, and Noam Yuchtman, “AI-tocracy,” Technical Re-
port, National Bureau of Economic Research 2021.
Besley, Timothy and Torsten Persson, “The origins of state capacity: Property rights, taxation,
and politics,” American economic review, 2009, 99 (4), 1218–44.
Bonney, Richard, The rise of the fiscal state in Europe c. 1200-1815, Clarendon Press, 1999.
Brecke, Peter, “Violent conflicts 1400 AD to the present in different regions of the world,” in
“meeting of the Peace Science Society” 1999, pp. 8–10.
Brewer, John, The Sinews of Power, Cambridge, Mass.: Harvard University Press, 1989.
Cantoni, Davide, Cathrin Mohr, and Matthias Weigand, “The rise of fiscal capacity,” Technical
Report, University of Munich 2022.
Coffman, D’Maris, Excise taxation and the origins of public debt, Springer, 2013.
21
Cox, Gary W, Marketing sovereign promises: Monopoly brokerage and the growth of the English state,
Cambridge University Press, 2016.
Dal Bó, Ernesto, Pablo Hernández-Lagos, and Sebastián Mazzuca, “The paradox of civilization:
Preinstitutional sources of security and prosperity,” American Political Science Review, 2022, 116
(1), 213–230.
Dickson, Peter George Muir, The Financial Revolution in England: a study in the development of public
credit, 1688-1756, Routledge, 2017.
Dincecco, Mark, Political transformations and public finances: Europe, 1650–1913, Cambridge Uni-
versity Press, 2011.
Findlay, Ronald and Kevin O’Rourke, Power and plenty: Trade, war, and the world economy in the
second millenium, Princeton University Press, 2007.
Gennaioli, Nicola and Hans-Joachim Voth, “State capacity and military conflict,” The Review of
Economic Studies, 2015, 82 (4), 1409–1448.
Glete, Jan, Navies and nations : warships, navies and state building in Europe and America, 1500-1860,
Stockholm: Almqvist & Wiksell International, 1993.
Heblich, Stephan, Stephen J Redding, and Hans-Joachim Voth, “Slavery and the British Indus-
trial Revolution,” Technical Report, National Bureau of Economic Research 2022.
Henriques, António and Nuno Pedro G Palma, “Comparative European Institutions and the
Little Divergence, 1385-1800,” Available at SSRN 3496613, 2019.
Hersh, Jonathan and Hans-Joachim Voth, “Sweet diversity: Colonial goods and the welfare gains
from global trade after 1492,” Explorations in Economic History, 2022, p. 101468.
Koyama, Mark and Jared Rubin, How the World Became Rich: The Historical Origins of Economic
Growth, John Wiley & Sons, 2022.
Levy, Jonathan, Ages of American Capitalism: A History of the United States, Random House, 2021.
Mintz, Sidney, Sweetness and power: The place of sugar in modern history, Penguin, 1985.
Mokyr, Joel, The Enlightened economy an economic history of Britain 1700-1850, Yale University Press,
2010.
Murphy, Anne L, “Demanding ‘credible commitment’: Public reactions to the failures of the early
financial revolution 1,” The Economic History Review, 2013, 66 (1), 178–197.
North, Douglass C and Barry R Weingast, “Constitutions and commitment: the evolution of in-
stitutions governing public choice in seventeenth-century England,” Journal of Economic History,
1989.
22
O’Brien, Patrick, “European Economic Development: The Contribution of the Periphery,” The
Economic History Review, 1982, 35 (1), 1–18.
, “The Political Economy of British Taxation, 1660-1815,” The Economic History Review, 1988, 41
(1), 1–32.
, “The nature and historical evolution of an exceptional fiscal state and its possible signifi-
cance for the precocious commercialization and industrialization of the British economy from
Cromwell to Nelson,” The Economic History Review, 2011, 64 (2), 408–446.
Palma, Nuno, “Sailing away from Malthus: intercontinental trade and European economic
growth, 1500–1800,” Cliometrica, 2016, 10 (2), 129–149.
Pomeranz, Kenneth, The great divergence: China, Europe, and the making of the modern world economy,
Princeton University Press, 2000.
Ramsey, Frank, “A Contribution to the Theory of Taxation,” The Economic Journal, 1927, 37
(March), 47–61.
Sánchez de la Sierra, Raúl, “On the origins of the state: Stationary bandits and taxation in Eastern
Congo,” Journal of Political Economy, 2020, 128 (1), 000–000.
Thomas, R. and N. Dimsdale, “A Millennium of UK Data,” 2017. Bank of England OBRA dataset.
Tilly, Charles et al., “Western state-making and theories of political transformation,” The formation
of national states in Western Europe, 1975, 638.
Vries, Jan De, “The industrial revolution and the industrious revolution,” The Journal of Economic
History, 1994, 54 (2), 249–270.
Williams, Eric, Capitalism and slavery, third edition. ed., Chapel Hill: University Of North Carolina
Press, 2021.
23
SUPPLEMENTARY APPENDIX, NOT FOR PUBLICATION
Table A.1: British Excise and Customs Revenues (in £1,000), 1689-1720
Supplementary Appendix — 1
Table A.2: British Excise and Customs Revenues (in £1,000), 1721-1757
Supplementary Appendix — 2
Table A.3: British Excise and Customs Revenues (in £1,000), 1758-1794
Supplementary Appendix — 3
Table A.4: British Excise and Customs Revenues (in £1,000), 1795-1823
Note: These tables report the yearly excise and customs revenues of the British government (in £1,000). Columns 1-3
report the revenues paid to the Exchequer collected by the excise, and columns 5-7 report the revenues paid to the
Exchequer collected by the customs. These exchequer revenues are net of management costs, drawbacks, and bounties
or other charges paid out of the revenues. Columns 2-3 for excise and columns 6-7 for customs disaggregate the
exchequer revenues into those collected from traded and domestic goods. Columns 4 and 8 report the revenues that
finance bounties or equivalent charges for national objectives. These revenues never reached the Exchequer and are
therefore not included in the totals of columns 1 or 5. The excise revenues paying for bounties (in col. 4) are solely
derived from traded goods. All the customs revenues paying for bounties (in col. 8) are collected from traded goods.
Cells are left empty whenever data is missing. See section 4 for a discusion of the sources and methodological choices.
Supplementary Appendix — 4
Figure A.1: Growth of British Government Debt in Times of War, 1691-1820
800
700
600 Napoleonic
Revolutionary Wars
Wars
500
400
300 American
Revolution
200 Seven
Years'
War
Austrian
100 Nine
Spanish
Succession
Years'
War Succession
0
1691 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1820
Year
Note: This figure plots British government debt in £1,000,000 over time, with major wars shaded in gray. See section 2
for a discussion of this figure. Source: Mitchell (1988).
Figure A.2: Excise Revenues by Taxed Good in 1756, Sample of CUST 145/22
Note: This figure reproduces a sample of the archival records organized under CUST 145/22. These records report
excise revenues by taxed good from 1788 forward. For earlier years (e.g. 1756 in this figure) the category of narrow
excise (cf. row 1 of this figure) needs to be further disaggregated. See section 4 for a more detailed discussion.
Supplementary Appendix — 5
Figure A.3: Excise Revenues by Government Act in 1750-55, Sample of CUST 145/12
Note: This figure reproduces a sample of the archival records organized under CUST 145/12. These records report
excise revenues by government act. See section 4 for a more detailed discussion.
Supplementary Appendix — 6
Figure A.4: Comparison of Aggregate Revenues with Mitchell (1988), with Bounties
Panel A: Excise Revenues
British excise revenue in £1,000,000
26
24
22
20
18
16
14
12
10
8
Mitchell (1988)
6
4
Our Data with Bounties
2
0
1689 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1823
Year
11
10
3 Mitchell (1988)
Our Data with Bounties
2
0
1689 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1818
Year
Note: This figure compares the aggregate patterns of excise and customs revenues of the British government (in
£1,000,000) as calculated by the authors with those reported in Mitchell (1988). Panel A reports the excise revenues
for 1689-1823, and Panel B reports the customs revenues for 1689-1818. The black, solid line plots revenues as paid into
the Exchequer plus revenues spent on bounties and other national objectives. This figure reproduces the plots of Figure
1, but with Exchequer plus bounty revenues instead of solely Exchequer revenues. Gaps in the lines indicate years with
missing data. We assume zero excise revenues spent on bounties before 1761. See section 4 for a description of the data
and methodology.
Supplementary Appendix — 7
Figure A.5: Revenues from Traded Goods and Domestic Goods, with Bounties
30
25
20
15
10
Traded with Bounties
Domestic with Bounties
5
0
1711 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1818
Year
Note: This figure decomposes the excise and customs revenue of the British government (in £1,000,000). The black
area plots the levels of revenues from traded goods, and the grey area plots the revenues from domestic goods. These
revenues are the sum of Exchequer revenues and revenues raised to pay for bounties and other national objectives,
as calculated by the authors. This figure reproduces the plots of Figure 2, Panel B, but with Exchequer plus bounty
revenues instead of solely Exchequer revenues. Years with missing data are linearly interpolated. We assume zero
excise revenues spent on bounties before 1761. The figure starts in 1711, as we do not observe data on bounties before
that year. See section 4 for a description of the data and methodology.
Supplementary Appendix — 8
Figure A.6: Revenues from Traded Goods as Share of Excise and Customs, with Bounties
.65
.6
.55
.5
.45
.35
.3
Mitchell (1988)
.25
.2
1689 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1818
Year
Note: This figure compares the share of British customs and excise revenue from traded goods as computed by the
authors with the share as reported in Brewer (1989) and using the data in Mitchell (1988). The grey, dashed line plots
the share following Brewer (1989) in treating customs revenue as coming from traded goods and excise as coming from
domestic production. The black, solid line plots the share of revenue from traded goods following the authors’ calcu-
lations and using disaggregated customs and excise data. Our Data with Bounties is the sum of Exchequer revenues
and revenues spent on bounties and other national objectives. This figure reproduces the plots of Figure 3, but with
Exchequer plus bounty revenues instead of solely Exchequer revenues. Gaps in the lines indicate years with missing
data. We assume zero excise revenues spent on bounties before 1761. The black line starts in 1711, as we do not observe
data on bounties before that year. See section 4 for a description of the data and methodology.
Supplementary Appendix — 9