Tải bản đầy đủ - 0trang
Putting the concept of 'purchasing power' into perspective
Statistical argument: construction, uses and controversies ...
at the same time, a complex one to quantify, since the conventions on what counts and on the
way in which it should be counted are problematic.
To sum up on this point, purchasing power is a subject that is both extremely vast and,
in some respects, reductive. It is vast in that it concerns national accounting, the concept and
measurement of the purchasing power of household disposable income and possibly other
national accounting aggregates, and even the distribution of value added107. It is reductive
since this concept implicitly limits purchasing power to its link with market consumption,
which is an important determinant of standard of living but excludes other expenditure
(investment in housing in particular) and the non-market aspects that also help to determine
quality of life, including access to public services.
1.2. Using the retail price index to assess purchasing power: necessary for indexation but
a necessarily imperfect convention
How are we to assess the evolution of purchasing power? One obvious determining
factor is the evolution of disposable income. However, we also need to take into account the
evolution of the prices of the products we wish to acquire, including consumer goods and
market services. Changes in prices are generally presented in the form of indices, which
summarise a number of very diverse changes, with the price of some products rising and
In order to illustrate the problematic nature of indices and justify their relevance despite
their inadequacies, here are a few comments taken from a real-life example.
Montpellier airport in the 1990s. A taxi.
'Could you take me to INSEE'S regional office?'
'Ah yes, the liars!'
'I beg your pardon?'
'But yes, at INSEE, they say that prices are stable but my water bill has gone up enormously'.
'Ah I see why you might think that. In fact, when they say that prices are stable, they should
say that some prices have risen, that others have fallen and that, overall, the increases and
decreases cancel each other out. It is the general price index that is stable, not the price of
every product. As far as the price of water is concerned, I don't know what's happened to the
price of water in Montpellier; it's true it's going up in many places. But you need to look
closely at your bill: if it's gone up, it may be because of the price of water but it may also be
that you used more than last year. So then it would be an increase in the volume of water you
used and not just an increase in the price. You would have had a higher water bill, possibly
because of the prices but also because you consumed more. You'd have to see… In any case,
my colleagues at INSEE are not liars: they are asked to summarise a complicated reality in a
simple and usable way; they are told to bring into the public arena a single price index based
These are important economic and social issues, but we will not dwell on them here.
A. Gély et al.
on millions of daily transactions; in any case, if they get it wrong, which can happen, it’s not
with the intention of deceiving us.’
Was the taxi driver convinced? It’s doubtful… But could he be convinced if, for
example, he had the impression he was finding it more and more difficult to make ends meet?
And particularly if the stability of INSEE’s price index was cited as a reason for not raising
his wages or the fares paid by his customers!
However that may be, constructing synthetic price indices, particularly consumer price
indices, is justified by the regulatory power they wield. The political, economic and social
implications of purchasing power are significant, since indexation affects the level and
evolution of the minimum wage and serves, or did serve, as a reference point for the uprating
of wages and other sources of income, such as social security benefits (e.g. pensions), the
RSA108 and even contractual arrangements such as maintenance allowances.
Thus the monitoring of prices and of the way they are evolving is a subject for both
debate and conventions. These conventions change because social relations, politics and
lifestyles themselves also change. Analysis of these conventions shows that questions related
to purchasing power were long based on a double convention. The first was that earnings
should evolve in line with prices. The second was that the evolution of prices should be
measured by the consumer price index (CPI), with all the history and conventions associated
The retail price index, which measures the change in the price of a basket of goods and
services, has been calculated and published by the French national statistics service since
1913. It was originally a response to the anger of the Parisian working classes, which was
triggered when the prices of basic foodstuffs and rents exploded in 1911. At that time, the
index was calculated only for Paris; it contained just 13 items (including soap and tallow for
candles) and was based on the consumption of manual and white-collar households with two
children. The index was expanded in the interwar period to include about 30 products. The
200 item barrier was crossed after the liberation of France in 1944 and robust consumer
surveys strengthened its representativeness in the 1970s, when the index became national.
This very brief outline of its history should not lead us to believe that the evolution of the
retail price index has been linear and unimpeded. On the contrary, its construction has been
chaotic and subject to pressures from all the parties involved in the indexation of contracts,
principally the state, the employers' associations and the manual workers' trade unions when
the sliding wage scale clause was adopted and constituted the rule for wage setting (between
1952 and the early 1980s). The following are some of the important moments in the
development of this index, and they show quite clearly that its history has been anything but
smooth: the manipulation of the price index at the end of the Fourth Republic, when
governments froze the prices of the public services included in the index and imported
massive quantities of the products included in it; the obligation imposed on INSEE between
1957 and 1962 to calculate and publish a minimum wage index (the so-called 179 item
RSA = revenu de solidarité active, which provides a minimum income for the unemployed and
Statistical argument: construction, uses and controversies ...
index); the dissemination of an alternative index calculated by the CGT in the 1970s and,
more recently, the pressures to exclude tobacco from the index.
2. Origin and emergence of the concept of purchasing power: a constantly debated
When speaking of developments over time, statisticians have long favoured the
expression ‘evolution of earnings in constant francs’ (and now ‘in constant euros’) over the
concept of ‘purchasing power’, preferring to reserve the latter term for comparisons between
different economic and monetary spaces, where price levels are not directly comparable (this
is what is known as ‘purchasing power parity’ or PPP).
In France, this notion of purchasing power does not seem to have established itself (in
these terms) in public debates until fairly recently.
More generally, the more distant past is approached in terms of consumer prices rather
than purchasing power (Touchelay 2013). However, while it is important (and we will return
to it later), the price index is only the denominator of the measure of purchasing power; it is
the numerator that reflects what is most crucial, since it is made up of the resources that
determine purchasing power. Moreover, a price index, by its very nature, can be used to
assess the evolution of purchasing power only, not its level.
Studies from the glory days of the Commissariat Général du Plan (the French State
Planning Commission) scarcely mention this terminology at all. For example, a detailed study
from the early 1970s that made extensive use of ‘social indicators’ does not mention it at all,
and purchasing power was not one of the policy objectives of the economic planning acts.
Other more specialist research centres were no more forthcoming. Thus the CERC (Centre
d'étude des revenus et des coûts), founded in 1966, did not begin systematically to present the
evolution of income in terms of ‘purchasing power’ until the 1980s. CERC disappeared in
1993, after years of only slight changes in purchasing power. Was this just chance? The final
example is INSEE’s flagship publication ‘Données sociales’, first published in 1971: although
‘purchasing power’ was certainly mentioned, it played only a token role that tended to shrink
further with each successive edition before it disappeared totally from the 1990s onwards.
2.1. The price index controversy of the 1970s
One particularly important moment in the development of the price index in France was
the lively debate that went on in the 1970s, when prices were rising particularly sharply, about
the index produced by INSEE; there was even an 'inside' contribution in the form of a
document disseminated by the INSEE branches of the CGT and CFDT trade union
federations entitled: 'Is the price index rigged?'109. These trade unions emphasised, quite
rightly, the importance of what is known as the 'quality effect': to what extent can a possible
Paper archives: the report by the INSEE trade unions entitled: L'indice des prix est-il truqué/Is the price index
rigged? (20 pages + annexes, Paris, February 1974); a report by the association DIES Pouvoir d'achat : du
mauvais usage des statistiques/Purchasing power: on the misuse of statistics 16 pages, October 1980; references
compiled by B. Touchelay.
A. Gély et al.
improvement in product quality be considered as a reduction in price? The INSEE trade
unions went so far as to call for the index not to be published. One would have expected to
find the expression 'purchasing power' in this document, particularly because the use of the
price index as a reference point in legal processes such as the indexation of contracts or court
decisions on such matters as maintenance allowances or in wage negotiations was not exactly
unknown to the document's authors. However, it is entirely absent from the document.
The controversy quickly extended beyond the confines of INSEE and the technical
departments most involved and was taken up by the trade union confederations. Thus the
CGT started collecting its own price data, from which it calculated its own index. The trade
union used this alternative indicator, which incidentally was more an expenditure index than a
price index, as an ad hoc 'statistical argument' in pay negotiations and as a basis for
demanding an increase in the national minimum wage. The CFDT favoured an approach
based on typical budgets, which the three main trade union confederations calculated from the
end of the Second World War onwards up until the 1970s. By calculating 'typical budgets', the
CFDT sought to show that workers could not live decently with earnings below the uprated
minimum wage for which they were calling.
These debates, which raged during the 1970s, tended to disappear from the public space
during the 1980s and 90s, for several reasons. Among them, the following are particularly
notable: disinflation (the annual rate of inflation in the French economy fell from more than
10% in the 1970s to about 2% in the 2000s); the de-indexation of wages in 1983; the fact that
the producers of the 'INSEE index' had to some extent taken account of the criticisms levelled
at it in the 1970s; the cost to the CGT of compiling its own index (collection of price data,
statistical analysis of the data, etc.) and the fact that in some months the CGT index increased
less rapidly than the INSEE index …
In the end, activists and the trade union confederations gradually stopped using the
alternative indices in wage negotiations. However, these debates were far from concluded.
2.2. The adoption of a 'tobacco-free' index in the 1990s
One important event, which is significant in more ways than one, took place in 1990.
This was the adoption as an instrument of indexation of a consumer price index (CPI) that
excluded tobacco. This political decision arose officially out of a concern with public health:
increasing the taxes on tobacco could both increase the state's revenues and reduce the
consumption of tobacco, whose harmful effects on health were well known. However, such an
increase was going to push the price index up and give rise to demands for wage increases in
order to maintain purchasing power. This is why it was decided to exclude tobacco from the
index. However, the statisticians declared that consumption should not be assessed morally
but objectively and pointed to the international definitions of consumption that France had
accepted; their arguments were strong enough for them to obtain agreement on continuing to
calculate and publish the full price index, including tobacco. Nevertheless, the legislature
made it compulsory to use a CPI that excluded tobacco for the purposes of indexation. This
regulation is still in force today.