. 3
( 11)


Parliamentary Coalitions

Tory Equilibrium
Landed anti-war Tories Monied anti-war Tories

Land Capital
Fiscal axis

Figure 2.3. Whig and Tory bargains in the British polity in the eighteenth century.

As long as the Whigs held a majority, then there was a natural pro-war
equilibrium in this “policy space.” I denote this point as the “Whig equi-
librium” in Figure 2.3.
By the end of the War of Spanish Succession, in 1713, debt had in-
creased from the 17 million sterling of 1697 to 36 million sterling (Brewer,
the dominance of Catholic France and be more willing to engage in war. In contrast,
many Tories (such as Henry St. John, Viscount Bolingbroke, the minister under Queen
Anne), had strong Stuart allegiances and might even be Jacobite (favoring James™s
son, Edward, for monarch after 1714). In general, then, Tories might dislike the
possibility of war with France. Such cleavages are re¬‚ected in the positioning of the
various factions in Figure 2.3.

Architects of Political Change

1988: 30). War weariness brought a Tory majority in the election of Oc-
tober 1710. Stasavage notes that interest rates on long-term government
debt jumped from 6 percent to 10 percent after the election, resulting
from the increased probability of default.8 The creation of the South Sea
Company in 1711 also re¬‚ected an attempt to force an involuntary reor-
ganization of government debt. Queen Anne, during her reign (1702“14),
had supported Tory preference for peace. Her ability to affect the compo-
sition of cabinets gave her the power to modify the political equilibrium.
A probable outcome in such a situation is labeled the “Tory equilibrium”
in Figure 2.3. After her death, the Hanoverian dynasty, which was more
aggressive against France, led to a period of Whig dominance.
However, the single Whig party eventually broke up into multiple fac-
tions. Without the king to act as a semi-collegium (and thus essential
to nearly every winning factional coalition) it is probable that the polity
would have descended into chaos. British and some American commen-
tators in the eighteenth century followed Montesquieu (1748) in arguing
that the combination of King, Lords, and Commons (representing monar-
chy, aristocracy, and democracy) formed a perfect constitution, combining
the best elements of each institution. The “balance” idea came originally
from Greek political theory. It is worth noting that eighteenth century
commentators might have read the history of Polybius on the Punic Wars
between Rome and Carthage 264 bc to 146 bc. Rome™s eventual suc-
cess was attributed to the balance of monarchy (the consuls), aristocracy
(the Senate) and democracy (the Popular Assemblies; see Goldsworthy,
2000, for discussion). Commentators would also have been aware of
Montesquieu™s argument that factionalism was an inherent feature of the
Roman polity, but that it was the great increase in Rome™s domain that
led to Rome™s collapse (Montesquieu, 1734; note that Kennedy™s imperial
argument is not dissimilar from Montesquieu™s).
It was this constitutional theory, mentioned by Hamilton, that pro-
vided the rationale for the balance of power between President, Senate,
and House after the Constitutional rati¬cation in the United States in
In fact, Montesquieu™s balance theory owed much to the in¬‚uence of
Viscount Bolingbroke, in exile in France after the Hanoverian, George I,
had come to the throne of the United Kingdom. Kramnick (1992 [1968])

The land tax, previously 20 percent or 4 shillings in the pound, was cut to 10 per-

cent. As discussed in Brewer (1988), this required an increase in the proportion of
government revenue raised by customs and excise.

Power and Social Choice

has used the term “the politics of nostalgia” to describe Bolingbroke™s
denunciation of the change in the nature of the political economic equi-
librium after the death of Anne in 1714. Since Bolingbroke™s argument
also impressed Thomas Jefferson later in the 1790s (Kramnick, 1990), I
offer an interpretation of its logic in terms of Figure 2.3.
As Figure 2.3 suggests, the Tories in Britain in the early eighteenth
century were less inclined to prosecute war with France, not just because
of “Jacobite” sympathies, but because of the cost and the seeming ne-
cessity of the land tax. During the War of Spanish Succession (1702“13),
escalating debt and interest payments had led to the creation of a stock
company”the South Sea Company. Essentially the company would bor-
row from commercial interests, lend to the government, and cover interest
with earnings from the Asiento concession trade with the Spanish colonies
of the Caribbean.
After the Hanoverian succession and the creation of the Whig majority,
the South Sea Company absorbed a greater share of government debt.
In January, 1720, Sunderland, ¬rst lord of the treasury, proposed that
the South Sea Company take over 30 million of the 51 million sterling
of government debt. The resulting share bubble collapsed in September
Robert Walpole stabilized con¬dence in the company by a transfer of
South Sea stock to the Bank of England and in April 1721 became both
Chancellor of the Exchequer and First Lord of the Treasury”“the ¬rst
prime minister in fact if not in name” (Williams, 1960: 179).
Walpole remained Chancellor until February 1742. The period of spec-
ulation in 1720 was seen by Bolingbroke to herald the end of the noble
constitution of England and the beginning of crass commercialism. Indeed
Bolingbroke™s fulmination against commercialism provided ammunition
to Thomas Jefferson when he decried similar speculations in the United
States in July 1791 when Hamilton™s First Bank of America opened. Bol-
ingbroke was correct to see the creation of the stable political economic
equilibrium in 1721 as the foundation of Britain™s commercial empire, but
he did not understand the reason for its stability. The danger for Walpole
in maintaining this equilibrium was that landed interests, whether Whig
or Tory, would repudiate the rapidly increasing debt of the government.
Walpole dealt with this problem by changing the composition of gov-
ernment revenue and by considerably increasing the amount raised by
customs and excise. It is not entirely accurate to call this situation an
equilibrium as this suggests immobility. Instead I conceive of it as a bal-
ance between differing political interests, stabilized by the preferences

Architects of Political Change

of the monarch. The resulting set of outcomes is denoted “Heart” in
Figure 2.3.
I contend that this device contributed to the stability of the political
heart and had a profound effect on Britain™s economic growth. Because
Britain™s imports were primarily foodstuffs and raw materials which were
relatively intensive in the factor of land, the restraint exercised on these
imports by the tariff had the effect of increasing the price of land, relative
to labor and capital (both of which were relatively abundant). It is compat-
ible with Rogowski™s (1989) analysis that landed interests, whether Whig
or Tory, would bene¬t from this contract initiated by Walpole. More-
over, because land became so valuable, it made sense to increase capital
investments in agriculture, thereby increasing productivity. This in turn
facilitated greater food production and made possible a rapid increase in
population. Because of greater intensity of the use of the factors of capital
and land in food production, labor might have been used less intensively.
Indeed there is some evidence that real agriculture wages declined in this
The increased value of the land could thus offset the tax on land, which
in turn was needed to ensure the credibility of payment of interest on gov-
ernment debt. Thus the safety of investment by the monied or commercial
interests in the government was guaranteed. By this method, Walpole se-
cured the stability of the contact between landed and monied interests.
This equilibrium was made even easier to maintain because Walpole was
able to sustain peace with France and Spain until October 1739. Obliged
by popular sentiment to declare war, Walpole ¬nally resigned in 1742.
To maintain this equilibrium from 1721 to 1739 Walpole constantly
bargained with the various factions of the Whig ascendancy (Namier,
1957; Brewer, 1976), bringing “placemen” and “stockjobbers” into the
Commons. Belief that these tactics corrupted the constitution was held
by many writers in addition to Bolingbroke. Sloan, for example, quotes
from Tobias Smollett™s History of England that

the Whigs . . . leaned for support on those who were enemies to the church and
monarchy, on the bank, and the monied interest . . . and prepared the minds of
man for slavery and corruption . . . They multiplied places and pensions, to in-
crease the number of their dependents.
(Sloan, 2001: 94)
See Allen (1988) and Crafts (1994). Note also that the “enclosures” of the Commons

in the eighteenth century may have been a further consequence. Porter has observed
that the “capitalist farm and the common ¬elds . . . became parables of industry and
idleness respectively” (Porter, 2000: 309).

Power and Social Choice

However, by increasing government revenue though the tariff, and in-
directly protecting land, Walpole reduced government debt from about
55 million sterling (with interest of 3.3 million) in 1721 to 47 million
(and interest of 2 million) in 1739 (Sloan, 2001: 301).
From the point of view of social choice theory, Walpole acted essen-
tially as a semi-collegium. Even though politics seemed highly factional-
ized and disordered, the corruption that Bolingbroke and Smollett decried
preserved the Whig equilibrium. The war in 1739 suggests that senti-
ment among the Whigs had moved up the war axis (as represented in
Figure 2.3).
From 1739 to 1784 (through three wars against the French, and lastly
against the American colonies), British government debt rose to 243 mil-
lion. This increase was primarily due to military expenditure in the Amer-
ican war, approximately 28 million in 1783 (Brewer, 1988: 30“9). The
land tax brought in only 20 percent of revenue, just over 2 million in
Because the principal source of revenue was customs and excise, this ne-
cessitated the creation of an extensive ¬scal bureaucracy (approximately
7,500 employees in 1770). Ferguson (2001) comments that this bureau-
cracy was roughly one person in 1,300. In France and Prussia, generally
considered highly centralized states, the comparable ratios were one in
4,100 and one in 38,000, respectively.
After Walpole, the general structure of the Whig equilibrium was
maintained, but factionalism became quite pronounced. I suggest that
some semblance of coherence was maintained because the Hanoverian
monarch, and particularly George III after 1760, came to play the role
of semi-collegium. However, because of the degree of heterogeneity in
the House of Commons and the House of Lords, actual policy could
lie within a larger domain, the set of outcomes denoted the Heart in
Figure 2.3.10

Suppose that I am correct in assuming the monarch acted as a semi-collegium, so

that the Nakamura number of the “constitutional game” in Britain was four. As note
5 asserts, chaos cannot occur if there are at most two dimensions of policy. Although
Figure 2.3 presents just two dimensions “ a land/labor axis and the war axis “ in
fact, land and capital represent two distinct dimensions. If there were indeed three
dimensions of policy, then there need not be an equilibrium as such. Instead, the
particular outcome would depend on bargains between the monarch and powerful
political factions. The “Heart” is intended to represent the domain within which
these bargains will reside (see Scho¬eld, 1999b for the de¬nition). The policy of the
British government toward the Colonies from 1763 to 1776 could indeed be called
inconsistent, rather than completely chaotic.

Architects of Political Change

A number of different inferences could be drawn, and were indeed
drawn by Hamilton, Madison, and Jefferson, from this British consti-
tutional experience. Hamilton inferred that a strong sovereign, or semi-
collegium, was necessary in order to overcome factionalism. Jefferson was
also acutely aware of the corruption that had attended the maintenance of
the Whig equilibrium and the evolution of Britain™s commercial empire.
Moreover, he understood that Britain™s expanding population required
the continual importation of food, primarily from the American colonies.
Control of trade, and the agricultural land in North America, was there-
fore a probable cause of con¬‚ict between Britain and its Atlantic colonies.
It would have been in Britain™s interest to control the price of imported
foodstuff, and, indirectly therefore, the price of land in North America.
Moreover, the customs and excise taxes levied in Britain were predom-
inantly regressive in nature, effectively taxing the poor rather than the
rich. This necessitated a degree of relative disenfranchisement, and possi-
bly accounts for the emphasis on virtual rather than direct representation
(Beer, 1993: 6). For these reasons the political economic equilibrium put
in place in the United States after Jefferson™s election in 1800 was very dif-
ferent from the British one. I contend that this U.S. equilibrium instituted
an agrarian empire.

2.7 the agrarian empire in north america
Chapter 3 makes the argument that the basis of the con¬‚ict between
Britain and its North American colonies from 1774 to 1783 was land.
After gaining Quebec and the entire region east of the Mississippi from
France during the Seven Years War (1756“63), Britain had to contend with
a ¬erce Indian revolt, often called “Pontiac™s War.” This was caused by
the ¬‚ow of settlers into the Ohio Valley. George Washington himself had
crossed the Cumberland Gap, had come into violent con¬‚ict with French
soldiers, and precipitated the war of 1756. Whether to avoid further con-
¬‚icts, or to defend Indian rights, the British attempted to restrict colonial
settlement by the Proclamation Act (1763) and then by the Quebec Act
(1774). This apparent act of tyranny infuriated the American elite, most
notably Franklin and Washington. In line with Riker™s (1964) theory of
the link between threat and constitutional change, Franklin negotiated
with the French for aid against the British. An argument presented in
Chapter 3 suggests that without French aid, revolution would have failed.
Promise of aid led to the Declaration of Independence in July 1776. In-
deed, French money, together with the role of the French Fleet and an army

Power and Social Choice

of 7,000 was crucial in the success of American arms against Cornwallis
in 1781 at Yorktown.
The war left the United States highly indebted. Beard (1913) mentions
a ¬gure of 50 million dollars (a fairly substantial ¬gure when compared
with the debts of Britain and France of 1 billion dollars and 600 million
dollars, respectively in 1784). Beard sees the rati¬cation debates of 1787
to 1788 as an argument over how to deal with this debt, with Hamil-
ton, and the Federalists, essentially proposing a hard-money principle on
the capital axis. However, the “land” axis was also important. As Riker
(1964) noted, the Spanish posed a threat on the Mississippi because of
their control of the Louisiana Territory and their declared intention to
block American exports through New Orleans. James Madison feared this
threat would destroy the fragile confederation of states, and it was this
fear that provided the federalist preference for union in 1787. However,
the extended domain of the proposed union implied, by Montesquieu™s
thesis on large republics, that factional11 collapse would occur. In a bril-
liant argument, ¬rst in “Vices of the Political System of the United States”
in April 1787, and later in Federalist X (November 22, 1787), Madison
reversed Montesquieu™s argument. Assuming that factions are generated
by common interests, he proposed that “the inconveniences of . . . States
contrary to prevailing Theory are in proportion not to the extent, but to
the narrowness of their limits” (Rakove, 1999: 79).
The heterogeneity of the extended Republic could then lead to reasoned
social choice. Chapter 3 suggests that this argument by Madison is similar
in logic to Condorcet™s Jury theorem of 1785.
Madison™s additional argument in Federalist LI was based on the the-
ory of balance of power, as derived from the British constitutional ideas
mentioned above. However, because the proposed constitution was fed-
eral in nature, bargaining between the central government and the states
would be more complicated than in Britain™s centralized system. We can,
perhaps, modify Figure 2.3 to study social choice in such a federal sys-
tem. First of all, note that there must be at least two dimensions of pol-
icy, land and capital. The “Hamiltonian” Federalists were concerned to
put in place a hard-money policy on the capital axis, so as to cope with
debt, and prepare the way for economic development. Opening up the
West to settlement would, however, be made easier with an easy credit

I interpret “factionalism” to mean the complete breakdown of any regularity to

government policy making, not the kind of inconsistency just mentioned.

Architects of Political Change

During the constitutional debates of 1787 and the Rati¬cation of 1788,
the argument revolved around differing interests of creditors and debtors
(as Beard has suggested), but also the necessity of having a centralized
authority to deal with foreign threats and with preventing the chaos of
factionalism. Madison™s argument in Federalist X suggests that the dif-
ferences in the House, Senate and President, together with the system of
cross-vetoes, would endow the polity with stability. House Representa-
tives, because of the small size of their constituencies, might be expected
to represent local interests. In contrast, the larger domain of interest of the
voters in each state would imply, by Madison™s heterogeneity argument,
that each Senator would encapsulate no particular interest, but rather the
concept of the public good appropriate to the state. Finally, the President,
chosen by majority rule from the electors of all states, would represent
some concept of the public good appropriate for the entire enfranchised
electorate. Because Congress could exercise a two-thirds veto against the
president, this implied that the President, together with over one-third
of both Houses, was collegial in those areas of policy controlled by the
Federal Government. Although this constitutional structure was intended
to avoid factionalism, it was not originally intended to deal with party
In fact, the interests of capital and land became extremely divided in the
United States, because of Hamilton™s attempt to recreate the Whig equi-
librium of the 1720s in Britain. To stabilize and fund the debt, Hamilton
proposed founding the Bank of the United States, backing it with receipts
from customs. However, unlike Britain, America™s imports were primarily
manufactured goods. In terms of Rogowski™s analysis, such a tariff would
support the relative price of the scarce factor”capital”more intensively
used in the production of these goods. Increasing the relative price of the
scarce factor is equivalent to reducing the relative price of the abundant
factor”land. Although Jefferson focused on the corruption attending the
attempt to create the bank, his other writings make it clear that he was
aware of the indirect effects of the tariff (I shall pursue this further in
Chapter 4).
In winning the elections of 1800 and 1804, Jefferson effectively con-
structed a coalition of the land, the Republicans, against the commercial
or mercantile interests of the Federalists. Because the Federalists were
keenly aware of the advantages of trade with Britain, the two parties
came to have distinct preferences on the land/capital axis as well as differ-
ing “international interests” in the long Napoleonic War between Britain
and France. The United States did, of course, eventually engage in war

Power and Social Choice

with Britain from 1812 to 1815, and by necessity had to institute a tariff
to ¬nance the considerable government debt that ensued. Nevertheless
the landed coalition put in place by Jefferson from 1801 to 1808 proved
fairly stable until the 1840s.
The basis for the Jefferson equilibrium was a compromise between
landed labor (small farmers in the West) and southern slave-owning
landed interests, at the expense, perhaps, of the commercial interests of
the Northeast, the Federalists, or the Whigs. The U.S. economy remained
focused primarily on raw material and food exports.12 As long as slave
interests did not threaten free labor, then the two parties, Whigs and
Democrats, had an intersectional (or north and south) constituency. Part
of the logic of the predominantly agrarian Democrat party was the ex-
pansion of this agrarian empire by conquest, ¬rst of the Floridas and then
of Spanish (or Mexican) domains in the West.
As Riker (1982) has discussed, maintenance of the Democratic coali-
tion became more dif¬cult, particularly as free labor and slave interests
came in con¬‚ict in the West. Moreover, the commercial interests of the
Northeast found that the dominance of the Democracy restricted eco-
nomic growth. Chapter 5 argues that the probable consequences of the
Dred Scott decision by the Supreme Court in 1857, and the debates of
Stephen Douglas and Abraham Lincoln in 1858, dramatized the threat
against free labor.
The compromise over slavery, inherent in the constitutional bargain of
1787, began to collapse, and the third dimension, the labor axis, be-
came increasingly important. Southern slavery interests followed Cal-
houn in arguing that an anti-slavery permanent (and hence tyrannical)
majority had formed. This violated the fundamental logic of Federal-
ist X and legitimated secession. In his inaugural speech of March 1861,
Lincoln asserted that the primary document of Union was the Declaration
of Independence, and this made secession illegitimate. Lincoln also re-
jected the “Crittenden” compromise, which would have granted all terri-
tory south of 36—¦ 30 to the slave interests. This seemingly made the Civil
War inevitable.13

Britain became heavily dependant on food supplies from the United States (in order

to feed its growing population) as well as cotton imports to maintain its textile
industry. In a sense, Britain and the United States maintained a symbiotic economic
relationship until the aftermath of the Civil War.
The social choice analysis presented above is given in terms of “winning,” vetoes,

etc., and is not strictly applicable to a situation of war. Nonetheless, the general logic
is valid. To maintain peace in the “three-dimensional space” required a compromise

Architects of Political Change


Republican point, 1860

Free agrarian interests Northern commercial Whigs

Jefferson point Hamiltonian point

Southern landed
Slave agrarian interests

Democracy slave interest


Figure 2.4. A schematic representation of land and capital in the United States, 1800“1860.

Figure 2.4 presents a schematic representation of the transformation
from the politics generated on the land/capital axis, circa 1800, to the
much more complex situation of 1860. The purpose of the ¬gure is to
show how the Republican/Federalist or Democrat/Whig splits were a con-
sequence of the suppression of the labor/slave axis. The threat from the
South transformed the underlying policy space, so that the intersection

between North and South. As discussed in Chapter 5, Lincoln himself put a ¬gure
of “two thousand million dollars” on the sum required to compensate the South for
the emancipation of slaves. Such a bargain was therefore impossible.

Power and Social Choice



Western Democrats


DOUGLAS, 29% Whigs

Liberal capital
BELL, 13%

Southern Democrats



Figure 2.5. A schematic representation of the election of 1860 in a two-dimensional policy

compromises collapsed. As Chapter 5 discusses in more detail, the 1860
election involved four presidential competitors”Lincoln, Bell, Breckin-
ridge, and Douglas. The point is made in chapter 5 that Lincoln, an ex-
treme risk taker, resolved the constitutional quandary over slavery. Figure
2.5 provides an illustration of the location of Presidential candidates in
the election of 1860.
Chapter 6 argues that, after the Civil War, and the resolution of the
slave dilemma, the policy space collapsed. The problem of the balance
between capital and land once again became paramount. The growth
of capitalism and manufacturing capability in the 1890s contributed to
the success of the hard-money Republican, McKinley, against his pop-
ulist Democrat challenger, William Jennings Bryan. In the 1930s, with the
Depression, the labor axis again became relevant. Over the period from
Roosevelt in the 1930s until Johnson in the 1960s, Democratic presiden-
tial candidates increasingly adopted policy positions that were liberal on
both the capital and labor axes. Simultaneously, Republican presidential
candidates adopted policy positions that were conservative on both axes.

Architects of Political Change

While it is common to assume that there is a single axis (“left”““right”)
in U.S. politics, this assumption seems to be valid only in particular in-
stances in U.S. history. In fact, the more general case may well be that
there are three distinct axes”land, capital, and labor. At the present time
(December, 2005), land, interpreted in terms of defense of the nation, is
clearly a separate issue. Indeed, there are already pronounced con¬‚icts
between concerns over civil rights (on the labor axis) and the need to
ensure surveillance and defense.
If this argument is correct, then theoretical inferences about political
convergence to an electoral center (Hotelling, 1929; Downs, 1957) appear
completely invalid. There is an additional point, originally made by Du-
verger (1954). Plurality rule (unlike proportional representation) forces
political parties to construct coalitions of differing interests. This can be
seen particularly in the election of 1860, where Lincoln constructed a
coalition of western agrarian interests and eastern commercial interests.
Because of the plurality rule of the electoral college, only 40 percent of the
voting population was suf¬cient to ensure victory. This suggests that when
constitutional quandaries beset the society, they can only be resolved by
the acts of political risk takers. In fact, the stronger hypothesis is that
plurality rule itself engenders risk taking to face fundamental democratic

2.8 the end of empire in britain
The same characteristic may hold true within the British parliamentary
system as well. To see this, consider Figure 2.6. I have suggested that the
third axis in British politics involved war, at least in the period until the
end of the Napoleonic con¬‚ict in 1815. Protection of land, via the tariff,
was the key element in the maintenance of the Whig equilibrium. How-
ever, after 1815, Britain™s commercial empire grew apace, and it became
increasingly dif¬cult to maintain the tariff against Whig interests. The
Irish famine, and the resulting concern of both Robert Peel and the Duke
of Wellington (in the Commons and the Lords) over civil unrest led to the
repeal of the Corn Laws (in May 1846). However, protection (particu-
larly with regard to wine from France) and high excise duties on alcohol
were both maintained. Nye (1991, 1992, 2006) contends that these were
calculated to maximize government revenue. After Peel™s betrayal of his
party, the Tories (or Conservatives) found it dif¬cult to secure a major-
ity against the Whig (or Liberal) free traders. In 1867, the Conservative™s
master tactician, Benjamin Disraeli, saw that the interest of land could be

Power and Social Choice


Many landed Whig Commercial Whig interests

Walpole Point

Free Trade Liberals
Tory Protectionism


Many landed Tory interests ( Few) commercial Tory

Figure 2.6. A schematic representation of land and capital in Britain, 1720“1846.

interpreted as “Empire,” and that this would appeal to a large proportion
of the British population. By the Reform Act of 1867, Disraeli completely
out-maneuvered the Liberal party leader, Gladstone, and almost doubled
the enfranchised population (Woodward, 1962: 187). In the election of
1874, the Conservatives gained 352 seats to the Liberals 242 (McLean,
2002: 92). By this time, the “war” axis of 1720 to 1815 had become
transformed into a religious axis. From 1886 there were two Irish fac-
tions, one Catholic and the other Liberal Unionist. The alliance of these
factions with the Conservative or Liberal parties made coalition politics
very dif¬cult until the Anglo-Irish Treaty of 1921 (McLean, 2002, ch. 7).
The rhetoric of empire generally served the Conservative Party well
until 1906. In the early 1900™s, Joseph Chamberlain tried to inte-
grate the issues of protection and empire, by his strategy of “Tariff

Architects of Political Change

Reform””“without preferential tariffs you will not keep the Empire”
(McLean, 2002: 122). While empire appealed to voters, tariffs did not.
The Conservative vote dropped from 50 percent in 1900 to 43 percent in
1906, and the Parliamentary seat share from 60 percent to 23 percent. At
the same time, of course, electoral reform and the increasing importance
of the labor axis led to the formation of the Labour Party (with 6 per-
cent of the vote and 5 percent of the seats in 1906; McLean, 2002: 88).
Over the decades of the twentieth century, the Conservatives kept their
pro-empire position, but moved to a pro-capital position, squeezing the
Liberal Party against Labour. Again, it is common to assume that British
politics is entirely determined by a labor/capital axis, with the Labour
Party at one end and the Conservative Party at the other.
However, empire has always been a component of the Conservative
Party rhetoric. In fact, the move by the Conservatives to a pro-capital
position may have only occurred during the prime ministerial terms of
Margaret Thatcher after 1979. Privatization of national industries and
the control of trade unions can, after all, be interpreted as pro-capital.
Her appeal to nationalism did prove electorally useful. However, after
Thatcher, the Conservatives have found that empire or “nationalism,”
when invoked against the EU, is dif¬cult to use as an electoral tactic. In
1997, the Conservatives took 31.4 percent of the popular vote (and 25
percent of the seats) against Labour™s 44.4 percent of the vote (and 63.6
percent of the seats). In 2001, these ¬gures were very little changed. Even
in the election of May 2005 (after Blair, by his actions on Iraq, caused
some proportion of the electorate to regard him with distrust), the Labour
Party vote share was still over 35 percent, while the Conservatives were
only able to raise their vote share to about 31 percent.
To show the relevance of the two dimensions (capital and nationalism)
in British politics, Figure 2.7 shows voter and party positions in Britain in
1997. Factor analysis of survey responses clearly showed two dimensions,
with two dense electoral domains”one pro-Europe, and the other, larger
one, fairly anti-Europe. Party positions were obtained from a survey of
Members of Parliament. Liberals were centrist economically and very pro-
Europe. The single Scottish Nationalist Party (SNP) member was much
less pro-Europe. Ulster Unionists (like Liberal Unionists in the past) were
more pro-labor than Conservatives, but basically opposed to European
The Duverger (1954) hypothesis, that two parties tend to dominate un-
der single member districts and plurality, is clearly not entirely valid in the

Power and Social Choice

PC Lib






’2 ’1 0 1 2

Economic Issues

Key: Lab = Labour, Cons = Conservative, Lib = Liberal and Social Democrat,
UU = Ulster Unionist, SNP = Scottish Nationalist, PC= Plaid Cymru.
Figure 2.7. Estimated voter distribution and party positions in Britain, 1997.
Source: Adapted from Scho¬eld and Sened, 2006.

British case. There is, however, a more general point to be made on the ba-
sis of this thumbnail sketch of British politics. Because there are generally
at least two axes of policy, plurality electoral rules force parties to con-
stantly adapt their position in response both to internal quandaries (e.g.,
reform, the Irish problem) and external ones (e.g., trade competition, war,
European union). Downsian convergence does not occur. Instead, politics
seem to exhibit a dynamic stability, where major policy shifts are en-
gineered by risk-preferring politicians (e.g., Peel, Disraeli, Chamberlain,
Lloyd George, Churchill, Thatcher). Riker (1986) coined the term heres-
thetic for the logic of these policy shifts. My inference from both United
States and British political history suggests that it is plurality rule itself,

Architects of Political Change

in both presidential and parliamentary systems, that engenders such risk-
raking heresthetic. It is possible, in fact, that plurality rule itself creates
dynamic stability and leads to “¬t” political choices.
The converse hypothesis is that PR systems lead to compromise, and
limit heresthetic and risk taking. This cannot, of course, be universally
valid, because Hitler rose to power under PR (Riker, 1953: 115; Evans,
2003: 259). However, it is possible to understand this latter phenomenon
as a response to extreme social uncertainty in the context of the Depres-
sion. This is one of the topics of Chapter 7.
This suggestion about the way democracy functions is, of course, con-
sistent with, but a development of, Condorcet™s so-called Jury Theorem
(Condorcet, 1785) and is a persistent theme of this book. Democracy, if
well constructed, leads neither to chaos nor equilibrium, but to constant
transformation, a kind of “structural stability.” This very general notion
is the topic of Chapter 8.

2.9 concluding remarks
In many respects Madison was the architect of change who propounded
the theoretical principles of the U.S. Constitution. These derived from
both the Scottish and French Enlightenments as represented by David
Hume, Adam Smith, and Condorcet.14 It is sometimes argued that the
two Madisonian principles of heterogeneity (in Federalist X) and balance
of power (in “Federalist LI”) are contradictory (Dahl, 1956, 1998, 2001;
McLean, 2003). From a social choice perspective, however, the two prin-
ciples are complementary. The essence of the Madisonian scheme at the
federal level is that it involves three parallel methods of aggregating pref-
erence. First, consider presidential elections. Initially, the electoral college
was chosen by state legislatures. Today, the electoral college is determined
by popular vote, but based on a system of weighting states. Generally,
states with low populations are advantaged, and this can lead to anoma-
lous outcomes, as in 2000. In general, however, the plurality nature of

Although the in¬‚uence of the Enlightenment on the ideas of Hamilton, Jefferson and

Madison has long been recognized (e.g., Commager, 1977), this book emphasizes
the speci¬c in¬‚uence of Condorcet on Madison and Jefferson. Chapter 4 discusses
the connection between the Jury Theorem of Condorcet and Madison™s arguments
in Federalist X, and suggests that the Jury Theorem can give a deep justi¬cation of
democracy. This theme is pursued with the more formal apparatus presented later
in Chapter 8.

Power and Social Choice

the election gives a clear winner. The election of Lincoln in 1860 with
40 percent of the popular vote, and 60 percent of the college, is an il-
lustration. Because the context is winner take all, it is intuitively clear
that presidential contests may tend to go to risk takers (e.g., Roosevelt,
Kennedy, Johnson). As the early discussion suggests, this may be an ad-
vantage in situations where war is possible.
In parallel, House and Senate elections aggregate preferences quite dif-
ferently. Since Senators are fewer in number, one might expect them to be
more risk averse than presidents, but more risk taking than representa-
tives. Finally, there are local elections for House and Senate at the state
level. This federal structure requires constant bargaining, and may be time
consuming, but it would seem to facilitate the ¬‚ow of information. The
key to the Constitution is the semi-collegial characteristic of the President
and the balancing of differing risk postures across political institutions.
The tentative interpretation offered here is that the Nakamura number is
high. This suggests that the polity can generally deal with complex prob-
lems except in exceptional circumstances, such as 1860, when extreme
heterogeneity destroys the political economic equilibrium.
The constitutional apparatus of the United States can be compared
with that of the EU. Because the electoral systems of the member states of
the EU are based on PR, it has seemed natural to rely on a weighted voting
system in the Council of Ministers. The Nakamura number of this system
can be computed (Scho¬eld, 1995). The larger nations, Germany and
France, for example, naturally have greater weight than small countries,
like Ireland. However, the system is essentially one of multiple vetoes.
This can have the effect of delaying decisions. Moreover, the device of
rotating the presidency of the EU means that there is no independent
It is well known that Condorcet relied on his Jury theorem for his
justi¬cation of a unicameral legislature, whereas Madison followed the
British Constitutional theorists in seeing the need for a parallel system
of representatives. As I have suggested here, the semi-collegial structure
of the British and U.S. polities in the eighteenth and nineteenth centuries
led to the stability of their political economic equilibria. In turn, these
contributed to the extension of the commercial British empire and the
agrarian U.S. empire. In the U.S. case, the violent con¬‚ict of the Civil War
seemed necessary to shift the Unites States toward a new equilibrium and
the creation of a commercial empire. Since Reconstruction, the U.S. polity
appears to have exhibited a degree of dynamic stability.

Architects of Political Change

Although the constitutional scheme of the Founders was designed to
escape the dangers of risk avoidance, chaos, and autocracy, one won-
ders about the future of U.S.“EU relations: how can a “hyperpower”
such as the United States with a risk-preferring president, negotiate with
a risk-averse confederation such as the EU? One purpose of this book
is to attempt to understand how the U.S. hyperpower came into being.
This may help us decide whether the United State is indeed an “imperial

Franklin and the War of Independence

3.1 introduction —
One way to understand the logic of the Declaration of Independence is
to attempt an estimate of the costs and bene¬ts and plausible subjective
probabilities associated with various outcomes that could follow from the
Declaration. While it is true that a general belief that “taxation without
representation is tyranny” gained ground in the colonies in the period
after the end of the Seven Years War in 1763, the studies that have been
carried out do not appear to give a realistic account of the motivations
of the British and American decision makers. In particular, in declaring
independence, the members of the Continental Congress expressed a will-
ingness to accept the great costs of war. To simplify greatly, each member
of Congress should rationally compute the expected costs of war after
such a declaration (say qC, where q is the subjective probability of war,
and C the subjective cost) against the expected costs of the status quo
(involving the probability that the British intended tyranny, T being the
cost of this tyranny). I give a more precise form of this calculation below.
Presented in this fashion, it is evident that unless the magnitude of T is
very high, then an “expected utility maximizer” would not choose inde-
pendence. An alternative way of interpreting this decision problem under
risk is to regard the absolute value of T as some prize to be gained by the
Colonies through a successful prosecution of a Revolutionary War. I ar-
gue that avoidance of taxation is simply inadequate as a suf¬cient prize.
In contrast, I emphasize that the British attempted to close the entire

This chapter is based on “Evolution of the Constitution,” The British Journal of Po-
litical Science 32 (2002): 1“23, by Norman Scho¬eld, with permission of Cambridge
University Press. Professor J. A. Pole, Timothy Groseclose, and Iain McLean kindly
made a number of very helpful comments on this earlier version of the chapter.

Architects of Political Change

Ohio Valley to settlement, by passing the Quebec Act in 1774. The Ohio
Valley can therefore be seen as the prize to the Colonists. Indeed the Act
exempli¬es British tyranny. This interpretation of the Quebec Act makes
intelligible the British actions. The British were themselves in a constitu-
tional quandary over the appropriate resolution of the con¬‚ict between
the Indian tribes of the Ohio Valley, and the colonial settlers. This con¬‚ict
had become obvious during the rebellion of many of the tribes, led by
Pontiac, from 1763 to 1765.
Chapter 2 brie¬‚y mentioned Riker™s (1964) argument that constitu-
tional change is induced by threat. From this viewpoint, the magnitude of
the prize of the Ohio Valley was a necessary cause or inducement toward
the creation of an independent federal republic. That is, if the problem
over the Ohio Valley had not been created, then there would have been
no declaration. This prize of the Ohio Valley was obviously not a suf¬-
cient cause because with some subjective probability, p, say, the Colonies
would lose a war with the British. I argue that any plausible estimation
of the probability of success against the British would be too low for an
expected utility maximizer to choose independence. However, I present
evidence to indicate that word of the promise of French aid reached the
Continental Congress prior to July 1776. This information would have
the effect of increasing the estimated probability of success. As a conse-
quence, an overwhelming majority of the Continental Congress came to
agree that a declaration of independence was rational. This secondary be-
lief in the rationality of such a choice can also be termed a core belief held
by members of the Continental Congress. By declaring independence, and
thus transmitting a signal to the colonial population, a more general core
belief in the justi¬cation of war was created.
The two instances of the 1776 Declaration and the 1787 Rati¬cation
(discussed in the next chapter) appear to pass through a number of sim-
ilar phases. As regards the Declaration, it is fairly evident that general
dissatisfaction with British actions became manifest, but the actual inten-
tions of the British must have been quite uncertain. It is reasonable to
call such a situation a quandary. After the passing of the Quebec Act,
uncertainty over British intentions diminished because the Act itself made
these intentions clear. We might interpret the situation in early 1776 as a
dilemma because expected costs of accepting the status quo and of declar-
ing independence were both high and comparable. The crucial informa-
tion about French aid transformed the dilemma into a choice, associated
with a core belief in the rationality of a declaration of independence.

Franklin and the War of Independence

Of course, it should be noted that this choice depended on the decision
by the French king, Louis XIV, to promise aid. As I mention later, had
Louis paid attention to his Finance Minister, Turgot, no aid would have
been provided. This cause was therefore somewhat contingent. However,
Benjamin Franklin had developed close contacts with the French court,
and, particularly with the Foreign Minister, Vergennes, and it is not un-
reasonable, therefore, to regard him, in some sense, as the architect of the
In a similar fashion, after 1784, the uncertainty over Spanish intentions
generated a quandary over appropriate actions. Once Spanish intentions
were declared through the attempt to close the Mississippi, the decision
problem became a dilemma. The costs both of accepting the status quo of
the Articles of Confederation, and of moving to a closer Union were high,
and again perceived to be comparable. As I suggest in the next chapter,
Madison™s arguments, about the ability of a republic to defend against
factionalism, transformed the dilemma into a choice.
These two transformations in the Constitution occurred rapidly.
Chapter 5 suggests that the Civil War may have occurred in a similar
fashion through a progression from quandary to dilemma to core belief
to choice.
These interpretations of constitutional change, while building to an
extent on Riker™s work, focus on somewhat different features. It is not
so much that Franklin and Madison engaged in heresthetic maneuvers
in 1776 and 1787. Both had a clear understanding of the nature of the
dilemmas facing the American people and acted through diplomacy and
argument to transform the beliefs of the Continental Congress and of the
Constitutional Convention.

3.2 the quandary of the declaration
of independence
The previous section suggested that the framework of fast or slow evo-
lution can be used to re-examine the causes of the War of Independence

Franklin was in Paris as Minister Plenipotentiary after 1776, and was very successful

in working with Vergennes to persuade Louis XVI that the war with Britain could be
won. For this reason we may see Franklin not only as the architect of the Declaration
of Independence, but the architect of the success of the War of Independence itself.
For the diplomatic background see Dull (1985). For Franklin in France see Hardman
(2000) and Schiff (2005).

Architects of Political Change

of 1776“83, as well as the origins of the Constitutional Rati¬cation af-
ter 1787. As I have noted above, the “cause” of the War can, of course,
be framed in terms of the belief that “taxation without representation is
tyranny,” and recent analyses have argued for the important role of such
“constitutional” ideas (see e.g., Rakove, 1996).
However, if we accept the outline presented above on the cause of a
belief cascade, then it is necessary to contrast the costs of this tyranny (T)
against the costs of war (C), weighed by the subjective probabilities, q,
that the British did intend tyranny; and p, that the Americans would lose
the war.
A serious problem for understanding the rational basis for the Decla-
ration is that Britain had shown itself, during the Seven Years War (also
called the French and Indian War) of 1756“63, to be an extremely pow-
erful military and ¬scal state (Anderson, 2000; Anderson and Clayton,
2005). Any reasonable calculation would give high values to p and C.
Even if there is some basis for believing that q was high (close to 1.0), but
T was low, then revolution would became irrational. Because there were
obvious differences of opinion in the Colonies over the rationality of rev-
olution, it is plausible that subjective estimates of T varied in the colonial
population. I contend, however, that many of the colonial elite had good
reason to believe T was high. Moreover, I argue that T had almost noth-
ing at all to do with taxation. In 1774, Britain passed the Quebec Act,
allocating control of the Ohio Valley to British Quebec. Many histori-
ans have commented on this Act in terms of British acceptance of French
Catholicism. The real effect for the Colonies was, however, noticed by
Thomas Paine:

Should affairs be patched up with Britain, and she remain the governing and
sovereign power of America . . . we shall deprive ourselves of the very means of
sinking the debt we have, or may contract. The value of the backlands which
some of the provinces are clandestinely deprived of, by the unjust extension of the
limits of Canada, valued only at ¬ve pounds sterling per hundred acres, amount
to upwards of twenty-¬ve millions, Pennsylvania currency; and the quit-rents . . .
up to two millions yearly.
(Foner, ed., 1995: 49)

Mark Egnal has also argued that

the revolutionary movement was led by an upper-class faction whose passion-
ate commitment to the rise of the New World was evident well before 1763.
The changes in property rights implied by the Quebec Act meant that the

Franklin and the War of Independence

expectations of many of the colonial elite (including Washington) of wealth from
land speculation were dashed.2

As a recent biographer of Washington has put it,
[T]he French Canadian fur traders were rewarded by the extension of the provin-
cial borders [of “French” Canada] to the Ohio River, thus wiping out all the
land grants not only of Washington, but of the Franklins and other land-hungry
It is interesting to note that Jefferson™s draft constitution for Virginia,
completed on June 13, 1776, makes no direct reference to the Quebec Act
but does include, in its list of denunciations of George III of Great Britain,
the assertion that the King endeavored “to bring on the inhabitants of our
frontiers the merciless Indian savages whose known rule of warfare is an
undistinguished destruction” (Peterson, ed., 1984: 337).
In the bill of indictment against George III drawn up by Jefferson in
late June 1776, there is a direct reference to the Quebec Act, that the King
gave his assent

for abolishing the free system of English Laws in a neighboring province, estab-
lishing therein an arbitrary government, and enlarging its™ [sic] boundaries, so
as to render it at once an example and ¬t instrument for introducing the same
absolute rule into these states.
(Peterson, ed., 1984: 21)

This phrase was retained in the Declaration of Independence of July 4.
I contend that this cost of tyranny was enormous, and almost counter-
balanced the likely costs of war. However, it was also crucial to the rational
calculations of the colonial elite that an ally be found. In fact, Louis XVI of
France agreed in March 1776, to help fund the Revolution to the extent of
ten million livres (more than $1.6 million). This was against the arguments
of the French Finance Minister, Turgot.

See Egnal (1988: xi), Baack (1998) and Phillips (1999: 18). Baack argues that the

fundamental cause of the Revolution was a mercantilist con¬‚ict between Britain and
the Colonies over extraction of “rents” from the Ohio Valley. From this perspective,
the cost, T, of British tyranny to the Colonies was enormous, but the game between
Britain and America was zero-sum, so the gain to Britain of this tyranny was also
enormous. I suggest later that the game, from the British side, was more complex.
Randall (1997: 264“5). Washington had, of course, been incensed by the earlier

Proclamation of 1763. See his letter to Charles Washington (January 31, 1770) in
Rhodehamel, ed. (1997: 134“7).

Architects of Political Change

Section 3.3 outlines a “game” played among the American Colonies,
Britain, France, and Spain, which underlay the Declaration. In fact, this
game is presented from the perspective of the Colonies. Some attempt is
also made to analyze the rationality of the British and French decisions.
However, questions can be raised about the British and French choices at
this point.
First of all, why did Britain pass the Quebec Act, which I argue is the
principal cause of the Revolution? I contend that the intention was not
primarily mercantilist, but was an attempt to protect the Indian tribes
of the Ohio Valley from the settlers. Pontiac™s Rebellion in 1763“5 had
made the depredations of the settlers on the Indians very clear to the
British. The Proclamation of 1763, closing the West, had little effect, and
the British tried to stem the ¬‚ood by a series of forts. It is possible that the
British attempt to defend the Indian tribes was due either to charity or to
gratitude for their assistance in the war with France from 1756 to 1763.
However, I consider it more plausible that the British perceived themselves
to be under a constitutional requirement to maintain the rule of law for
all subjects of the King. From this perspective, the Indians were being
deprived of their rights. This interpretation suggests that the British were
in a constitutional quandary, which they could only solve by threatening
force against the recalcitrant Colonies.4
This issue of Indian rights is worth pursuing further. During the Indian
rebellion under Pontiac all but three of the British forts in the Northwest
Territory had been taken prior to June 1763. The Proclamation of 1763
was meant to acknowledge Indian rights and help bring about peace.
The Proclamation line essentially de¬ned a huge area between the Ap-
palachians and Mississippi as Indian. The later treaty of 1765 with Pon-
tiac stipulated that, though the Forts were British, the land was not. As
a recent commentator has put it: “The . . . Crown . . . wanted to prevent

Beer has discussed the differences between the notion of virtual representation held

in Britain and the notion of actual representation, as it was to be articulated in
the Colonies, particularly by Madison. Under virtual representation, a member of
the House of Parliament was regarded as speaking on behalf of the whole nation.
He should shun “dependence” on the electors and follow his own judgment of the
public good. The Colonies™ rejection of this notion is part of the explanation for
the Revolution. But, of course, the welfare of the Indian tribes should be considered
under the public good, if they were perceived as subjects of the King. Thus, the notion
of virtual representation was, in a sense, doubly insulting to the Colonies. For these
concepts of representation, see Beer (1993: 164“6). For the activities of the British
Indian Department from 1763“76, see Colloway (1995). For details on Pontiac™s
Rebellion, see Anderson (2000: 535“53) and Jennings (1988, 2000).

Franklin and the War of Independence

˜unjust Settlement and fraudulent Purchase™ of Indian lands. . . . A question
still being argued is whether the Proclamation recognized a pre-existing
title or created it” (see Dickason, 1992: 188; see also Stagg, 1981a,b).
As Dickason (1992) notes, the legal questions were still being discussed
more than a century later in Canada. Indeed, in 1885 the judgment of the
Chancellor of Ontario was that the Indians had become “invested with a
legally recognized tenure of de¬ned lands; in which they have a present
right as to the exclusive and absolute usufruct, and a potential right of
becoming individual owners in fee after enfranchisement” (188).
I shall argue that the Declaration would have been irrational on the
part of the members of the Continental Congress had they not known
of the promise of French assistance. There is no direct evidence that
the Congress knew of the promise of French aid prior to July 4, 1776
because there are no existing memoranda of the Committee of Secret
Correspondence (which, under Franklin, negotiated with France). How-
ever, Benjamin Franklin™s papers include a letter, in code, from Charles-
Guillaume-Frederic Dumas (in Utrecht) to the Committee, dated April 30,
1776 (Wilcox, ed., 1982: 403“12). Because the company, Roderique Hor-
talez, set up to supply the Americans from France, was based in Utrecht,
it is a plausible inference that Franklin knew of the promise of 10 million
livres of French aid prior to July 4, 1776.5
Whether it was rational for France to provide this aid is dif¬cult to
determine. Turgot tried to dissuade the King because he foresaw the ¬s-
cal problems for France that would follow.6 In fact, economic historians
tend to agree that the eventual cost to France for the support it provided
the Colonies was in excess of 1 billion livres. The resulting debt precip-
itated the French bankruptcy of 1788 and, thus, the French Revolution
(Norberg, 1994: 252“98).
Initially, of course, Louis XVI must have believed that, by providing
aid to the Colonies, he would facilitate Britain™s losing the jewel of its
empire. However, he was careful to disguise the existence of the aid, and
formal treaties of friendship, commerce, and alliance with the Americans
were not made until February 1778. These were approved by the U.S.
Congress in May 1778, and by June of that year France and Britain were
at war.7 The aid provided in 1776 was, however, known to the British

The full story of this secret aid, and the involvement of Pierre-Augustin Caron de

Beaumarchais, has yet to be written.
See Roche (1998) for Turgot™s reform efforts prior to his dismissal in mid-1776.

Franklin was in Paris as Minister Plenipotentiary and was very successful in working

with Vergennes to persuade Louis XVI that the war with Britain could be won.

Architects of Political Change

Cabinet through the ef¬cient British spy network, and it was mutually
convenient for both Britain and France to maintain the pretence of peace
for two years after 1776. In France™s case, it was a question of waiting
to see whether the Revolution would succeed. If it did succeed, then the
potential gain to France would be the great opportunity, again, to open
up the Louisiana Territory. However, the ¬rst two years of the Revolution
were dif¬cult for the Colonies, and French aid was crucial (McCullough,
2005; Weintraub, 2005). Once the public decision to aid the Colonies
was made, then it would be rational to continue the aid in the form of
men and mat´ riel, in the hope of eventual success. We can infer that Tur-
got believed the gamble was too dangerous, because the risk was French
Once Louis XVI made the decision in 1776 to aid the colonies, reason-
able subjective estimates by the members of the Continental Congress of
the likely expected costs of war compared with acquiescence to the British
led to a majority decision to declare independence. To use the terminol-
ogy proposed here, a core belief was created, within the colonial elite,
that war could be justi¬ed in terms of expected utility calculations. If a
majority of the Continental Congress had not accepted this belief, then in-
dependence would not have been declared. Irrespective of the beliefs of the
population as a whole, there would have been no War of Independence in
The interpretation presented here of the American Revolution is that
it was triggered by a change in beliefs by (a majority of) the colonial elite.
This change was necessary for the Revolution; otherwise there would
have been no Declaration of Independence. The change, however, was
certainly not suf¬cient; changes in the beliefs of the colonial popula-
tion were also required. Historians have suggested various causes based
on religion or class for these general responses.8 In contrast, I have as-
serted that the reason for this elite belief cascade was the combination of
the passing of the Quebec Act and Louis XVI™s choice to provide aid. The
¬rst of these causes”the passing of the Quebec Act”is intelligible in terms
of British Constitutional understanding of virtual representation. The

For this reason we may see Franklin not only as the architect of the Declaration of
Independence, but the architect of Independence itself. See Stourzh (1954), Hardman
(2000 ) and Schiff (2005).
For example, Phillips (1999) emphasizes the common threat of religious antagonism

connecting the British Civil War, the American Revolution, and the American Civil
War. See also Draper (1996).

Franklin and the War of Independence

second cause”the decision by Louis XVI”is only weakly acknowl-
edged by historians. And the likelihood that Franklin knew of the
promise of French aid prior to July 4, 1776 has not been considered
An earlier account of Franco-American negotiations by Bemis (1935)
noted that a courier was dispatched from France to the Colonies in May
1776, with the news of the promise of aid. However, the ship was sunk,
and the information did not reach Philadelphia until December 1, 1776. It
is highly implausible that only one courier was dispatched. Because the sea
voyage was about two months, it is reasonable to infer that the promise
of aid did in fact arrive at the end of June 1776. I suggest that this promise
of French aid was pivotal in the sense of making the Declaration rational.
Of course, this cause was a highly contigent, unpredictable event.
Although French amity was crucial to the success of the revolution, re-
lations between France and the United States deteriorated after the French
Revolution. Indeed, during the administration of John Adams (Ferling,
1992; Ellis, 1993; McCullogh, 2001), from 1797 to 1800, France and
the United States were almost at war. In August 1800, by a secret treaty,
Napolean Bonaparte forced Spain to return Louisiana to France (Adams,
1986 [1889]: 247). Francois Toussaint had led St. Domingue (or Hispan-
iola) in revolt against French rule in 1791, and Bonaparte determined to
regain the island, and to add Louisiana to his dominion. To this end, he
dispatched a ¬‚eet and an army of twenty-eight thousand men under his
brother-in-law, Leclerc. As Menig (1986) has put it, by 1801 the “recon-
quest of St. Domingue, the return of Louisiana, and the recreation of a
great French American Empire suddenly became speci¬c and seemingly
speci¬c goals.” (337). President Jefferson was well aware of this threat,
particularly to New Orleans, and wrote to Robert Livingtone, his Min-
ister in France, that “[T]he impetuosity of [France™s] temper . . . render[s]
it impossible that France and the United States can continue long friends.
From that moment we must marry ourselves to the British ¬‚eet and na-
tion.” (Peterson, 1984: 1105).
By September, 1802, however, almost the entire French army was lost
to disease. Adams (1986 [1889]) argued that this defeat was the reason
that Napolean sold Louisiana to Jefferson. The purchase was negotiated
in New Orleans in December, 1803, and on March, 9 1804, Meriwether
Lewis witnessed the signing of the documents in St. Louis, thus capping
the “great gamble” that was made, ¬rst by Franklin against Britain and
then by Jefferson, against France (Cerami, 2003).

Architects of Political Change

3.3 the decision to declare independence
One way to deal formally with the Constitutional Congress™ decision in
July 1776 is to construct a game involving America, Britain, France, and
Spain. To do this however, in a technical fashion, requires estimating pref-
erences and Bayesian priors for the combatants. Instead, this section will
consider the problem in the context of decision making under risk. That
is to say, I consider the problem from the point of view of an “average”
member of the Congress, estimating the likely expected costs of choosing
to declare independence. Clearly, this estimate will vary among the mem-
bers of the Congress. However, this problem can be interpreted formally
in terms of what we may call a “Condorcet Jury,”and the rational choice
of an average “juror” will give insight into the choice of the Congress.
To emphasize the point again, it is not so much that the members of this
congressional jury do or do not prefer to declare independence. Rather it
is a question of whether each one believes that it is in the Colonies™ best
interests to declare independence.
First, consider the military and ¬scal capability of Britain prior to 1776.
As North and Weingast (1989) have shown, the Glorious Revolution of
1688 induced a constitutional and ¬scal revolution that better equipped
Britain to borrow and to ¬ght wars. As discussed in Chapter 2, Britain
increased its debt to about 120 million sterling during the various wars
of the ¬rst half of the eighteenth century, and yet was able to fund an
increasing proportion of this debt (Brewer, 1988).
During the Seven Years War, Britain demonstrated that it could solve
the logistic problem of arranging for a convoy of troop ships, together with
their protecting men-of-war, to arrive together and surprise the enemy. By
this means, Wolfe captured Quebec City in September 1759, and British
forces took Havana, Cuba, and Manila in the Philippines (on August 13
and October 7, 1762, respectively). The Havana force comprised fourteen
thousand men, ten thousand of whom became sick or died of tropical
disease (Pocock, 1998).
Perhaps in an attempt to bluff the Colonies out of rebellion, a British
¬‚eet of more than 130 ships under General William Howe arrived at the
port of New York by June 30, 1776 (Bobrick, 1997: 206). By July 12,
thirty-two thousand troops were in position to disembark from more
than 170 transports, protected by twenty frigates and ten ships of the
line. By August, the British naval presence was more than four hundred
transports and ¬fty ships of war. In the battle for New York in August
1776, Washington had fewer than three thousand men. With these facts in

Franklin and the War of Independence

mind, consider the decision problem facing the members of the Continen-
tal Congress in the month prior to July 4, 1776. Again, the probabilities
involved are q (the British will attack if independence is declared) and p (if
they attack, the Colonies will lose). The costs involved are T (of tyranny)
and C (of war).
Because war is a risky venture, we should incorporate the risk posture
of the protagonists. To keep the calculations as simple as possible however,
I shall ignore risk posture and deal only with expectations. Let us assume
that the purpose of the Declaration of Independence is to avoid the costs
of British tyranny. To further simplify, suppose that the prize of freedom
is T dollars. We can perform the calculations in Spanish silver dollars,
each one of which is worth (at that time) approximately six French livres.
A pound sterling was worth about four Spanish dollars. Total French
grants and aid in the initial stages of the war totaled 36 million livres, or
6 million dollars. For purposes of illustration, let us assume that the cost of
war to the Colonies, if the British attacked, could be estimated at C = 12
million dollars (excluding French aid). If the Colonies lose the war, the
cost is C + , where is the additional cost of defeat. For example,
the members of the Constitutional Congress could expect to be hanged
for treason after defeat. Appropriation of property, and so forth, after
defeat would suggest an order of magnitude for approximately equal
to C.
If the Colonies declare independence, and the British attack (with prob-
ability q), the Colonies win the war with probability (1 ’ p). Thus, the
expected utility is
q[(1 ’ p)(T ’ C) + p(’C ’ )]. (3.1)
If the British acquiesce, then the prize T is won with probability (1 ’ q).
Independence is a rational choice if and only if
q[(1 ’ p)(T ’ C) + p(’C ’ )] + (1 ’ q)T > 0. (3.2)
Rearranging terms gives the condition
(1 ’ qp)T > q(C + p ). (3.3)
For example, suppose beliefs are optimistic in the sense that p = q = 1 ,
and = 0. Then, the condition is T > 2 C.3
Middlekauff (1982: 150) suggests that the ¬gure for the cost of tax-
ation to the Colonies was 40,000 dollars or T = 200,000 dollars, if we
take this sum discounted over a decade. Chown (1994: 218) suggests a
¬gure of 16,000 pounds sterling or 64,000 dollars per annum, or perhaps

Architects of Political Change

320,000 dollars discounted over the decade. With C = 12 million dol-
lars, the cost (of avoiding taxation) could not possibly make rational the
expected costs of war. If the French do offer aid, then the cost of war
is divided between the French and the Colonies, so the cost facing the
Colonies is reduced to C . Moreover, French aid can be expected to in-
crease the probability q (that the British attack) to nearly 1. Suppose
French aid further affects the probable outcome of war by changing the
probability of defeat from p = 1 to p = 1 , say. Then, the requirement is
2 4
T > 4 C . This is only slightly different from the previous calculation, so
again, avoidance of taxation is not a suf¬cient cause of a declaration.
However, as noted above, Thomas Paine argued that the cost of tyranny
was the loss of the Ohio Valley, which he put at a value of 25 million
For an optimist ( p = q = 1 ), the calculation, assuming no French aid
and C = = 12 million dollars, becomes T > 2 (12 + 6) = 12 million
Given a prize of this magnitude, such an optimist would certainly de-
clare independence. A more prudent member of Congress might reason
that p = q = 0.9, in which case the requirement becomes T > 0.19 (22.8)
or T > 108 million dollars. For such an agent, a declaration (without the
promise of French aid) is not justi¬ed.
Suppose, however, an offer of French aid has been received, and the
prudent calculator reasons that the probability of defeat will, as a con-
sequence, fall from p = 0.9 to p = 0.5. With q = 0.9, the calculation
becomes T > 1’0.45 = 29.45 million dollars.
Thus, a risk-averse calculator, unwilling to accept high costs such as
C and , could reason that the value of the prize is close to the expected
costs. As indicated, the calculation depends on the amount of French aid,
the effect this will have on the prospect of victory, and on the likely costs,
, of defeat.
Some of the members of Congress, and of the population at large,
could be highly optimistic and believe in the rationality of a declaration,
even without the offer of aid. A more prudent calculator could reason
that not only would the costs of war be reduced by French aid, but that
probable French military assistance could directly increase the probability
of victory.
The implicit argument here is that variation in beliefs and risk pos-
tures among the members of Congress would lead to different calculations
about the rationality of a declaration of independence. Prior to news of
French aid, a majority would not have chosen independence. After the

Franklin and the War of Independence

news was received, however, even relatively risk-averse members could
choose independence, thus creating a core belief in the rationality of in-
From the point of view of the French, it would be rational to offer a
“pivotal” amount of aid, suf¬cient to create a majority in Congress in
favor of independence. The French offer, by changing beliefs as regards
p, and by funding some of the costs of the war, clearly would increase
the number in Congress favoring independence. A signal from the French
prior to July 4, 1776, was, however, crucial for this transformation.
After France entered the War in 1778, and with Spain as its ally to help
defend the French empire in the Caribbean, the ¬‚ow of men and mat´ riel e
rapidly increased. The crucial battle of the war”Yorktown”was won
through French assistance. In fact, a French ¬‚eet from Rhode Island and
one under Comte de Grasse, from the Caribbean, performed the British
trick of a coordinated arrival at the York River in September 1781. A
French army of 6,500 helped Washington lay siege to the British forces
under Cornwallis at Yorktown, while the French ¬‚eet stopped the British
ships from relieving the British army.
In mid-1776, it could not be known with certainty that French forces
would be supplied in this fashion. However, it was a rational hope that
Louis XVI would continue to help, and that this assistance would at least
exhaust the will of the British to continue the war.
On the British side, the logic of the problem involves the probability (r )
that the Americans will acquiesce if the British attack, and the probability
that the British will win ( p) if the Americans do not acquiesce.
Using B1 for “attack” and B2 for “do not attack,” and using u for
expected utility we obtain
u(B1 ) = (1 ’ r )[ p(T ’ C) ’ (1 ’ p)C] + r (T ’ C ), (3.4)
while u(B2 ) = 0.
Here, C is the initial cost of attack, due to placing a ¬‚eet in New York
in 1776 (say 100,000 dollars). C, as before, is the cost of full-scale war.
First of all, let us take T = T0 to be the “bene¬ts” of taxation, or 200,000
dollars. Obviously, if the British do not attack (B2 ), these bene¬ts are lost
through independence. Even with the optimistic value of p = 0.9 (for a
British victory), the condition requires that r is almost 1.0.
Thus, if the British have no knowledge of French aid to the Colonies,
then they should attack only if they expect the Americans to calculate that
victory is impossible and to acquiesce to the display of force. This assertion
depends on the calculation that British bene¬ts simply had to do with the



. 3
( 11)