1. Marshalling A Contradiction
In the third edition of his Principles, Marshall defended his consumer surplus analysis,
acknowledging that, in strict theoretical terms, it required that the marginal utility of
money remained constant. Provided, however, that there were only small changes in
"the neighborhood of the customary price" and that special care was taken in dealing
with "necessaries", the "substance" of the analysis remained valid and there would be
"very few practical problems" in which changes in the marginal utility of money would
have "any importance" [Marshall 1895, pp.208-9]. To illustrate the last point, he
argued that
5
For biographical information on Giffen, see Mason 1989, ch. 3.
6
Following George Stigler [1947; 1948], this point is generally accepted by historians of
economics. The news has, however, been slow to reach some quarters: “[A Giffen good is
named] after the nineteenth - century economist who first noted the possibility” [Varian 2010,
p.105].
4
There are however some exceptions. For instance, as Mr Giffen has pointed
out, a rise in the price of bread makes so large a drain on the resources of the
poorer labouring families and raises so much the marginal utility of money to
them, that they are forced to curtail their consumption of meat and the more
expensive farinaceous foods: and, bread being still the cheapest food which
they can get and will take, they consume more, and not less of it. But such
cases are rare; when they are met with, they must be treated separately [ibid.
p. 208].7
Contrary to the assumption in much of the subsequent literature, Marshall was not
concerned with explaining an upward-sloping market demand curve in the
Principles.
8
Rather, the point of his analysis was to defend the concept of consumer
surplus against the criticism that it was of little practical use because it was not
possible to provide accurate estimates of aggregate consumer material welfare if
there were marked differences in the distribution of income and thus the marginal
utility of money between different groups of transactors. Rejecting that criticism,
Marshall argued that the extreme case of the labouring poor increasing their
consumption of bread as the price rose was atypical. Such behaviour did not threaten
the estimation of consumer surplus with a downward-sloping market demand curve
because ‘such cases are rare’. To argue that the market demand curve was upward
sloping would have destroyed the point of his argument. In the Principles, the
question was not: what are the conditions in which a market demand curve has an
upward slope? Rather the question was: is the concept of consumer surplus of
practical use in estimating changes in consumer welfare, given a downward-sloping
market demand curve?9
7 The paragraph was largely unchanged in subsequent editions, although Giffen was elevated
to Sir R. Giffen (he was knighted in 1895) and “they must be treated separately” became “each
must be treated on its own merits" [Marshall 1961, 1, p. 132].
8
The following paragraphs draw on White 1990; 1991.
9
In the second edition of the Principles, possibly responding to criticism of the first edition,
Marshall had distinguished between an explanation for a positive relation between price and
quantity demanded in terms of a movement along, and a shift in, a market demand curve.
There was an "implicit condition in this law [of diminishing marginal utility] … It is that we do
not suppose time to be allowed for any alteration in the character or tastes of the man himself.
It is therefore no exception to the law that the more good music a man hears, the stronger in
his taste for it likely to become; that avarice and ambition are often insatiable; or that the
5
Marshall then produced a different argument in his Memorandum, which was first
written in 1903 and published in 1908 [Marshall 1926 (1908), pp.380 – 85]. The
market demand curve for bread was now claimed to be upward sloping, at least over
a range. With an increase in the price of wheat following the implementation of an
import duty, the "total demand" of “nearly the whole of the English people" for bread
would increase:
as Sir R. Giffen seems to have been the first to observe, a rise in the price of
wheat still leaves bread the cheapest food, which they will consent to eat in any
quantity; so that, having to curtail their purchases of more expensive foods,
they buy not less bread than they would have done, but more" [Marshall 1926
(1908), p. 382].
In a subsequent exchange of letters with a sceptical F.Y. Edgeworth, Marshall
defended his argument by referring to his own observations and to his familiarity with
statistics underpinning estimates of demand and supply elasticities for wheat
[Whitaker 1996, 220-224].10 He also referred to a hypothetical case to support his
argument, although that was not relevant for the argument about elasticity estimates
[Dooley 1992]. The key point, however, is that, within the space of eight years,
Marshall claimed that the English market demand curve for bread was both upward
and downward sloping.
virtue or cleanliness and the vice of drunkenness alike grow on what they feed upon"
[Marshall 1961, 1, p.94].
10 In the Memorandum, for example, when referring to "distant wheat-fields", Marshall noted
that "there has been a rise [in price], as Mr. Powers has well shown, even when the price in
Liverpool was falling rapidly in consequence of the great diminution in transport charges on
the west of the Mississippi, as well as between Chicago and Liverpool" [Marshall (1908) 1926,
p.383]. Peter Groenewegen has informed me that Marshall was referring to Henry Huntington
Powers (1859 - 1936), associated with the University of California, author of an article on
“Expansion and Protection” in the Quarterly Journal of Economics [Powers 1899], contributor
of three articles to the 1926 Palgrave’s Dictionary of Political Economy and author of the
following books: Lectures on Money (1893), Wealth and Welfare (1899); America among the
Nations (1918); and, America and Britain (1918). Marshall also cited Powers on US wheat
production in Industry and Trade [Marshall 1920, p. 776].
6
An explanation for Marshall's switch can be found in the different contexts in which
the two arguments were produced. In the Principles, he was attempting to confute
the critics of his consumer surplus analysis, particularly J.S. Nicholson, his former
student who was professor of political economy at the University of Edinburgh. In the
Memorandum, however, he was attacking the proponents, especially W.A.S. Hewins,
first director of the London School of Economics, of the use of tariffs within a system
of imperial preference.11 The question Marshall posed in the relevant section of the
Memorandum was: what will be the effect on the price of bread of imposing a tariff on
imported grain? The upward-sloping demand curve supported the claim that
domestic consumers, rather than foreign suppliers, would bear at least the full
incidence of the tariff.
The figure of Marshall as the origin of a single ‘Giffen conjecture’ erases the context,
contradictions, complexity and specific arguments of Marshall's texts, imposing a
question that, in the case of the Principles, would have destroyed the point of his
analysis. Nevertheless, it became common to refer only to the Principles in
subsequent references to Marshall having discussed a demand curve with a positive
slope. An early example was W.E. Johnson's discussion of demand curves in 1913,
which included perhaps the first published reference to “Giffen’s paradox” [Johnson
1913, p.484]. Despite subsequent claims to the contrary, Marshall did not use the
latter phrase, which is not surprising as he rejected the notion of a paradox with
regard to Giffen behaviour [Whitaker 1996, pp.222-23].12