AGGREGATION AND THE DISTRIBUTION OF INCOME
| DOI | http://doi.org/10.1111/j.1468-0084.1975.mp37002002.x |
| Date | 01 May 1975 |
| Published date | 01 May 1975 |
AGGREGATION AND THE DISTRIBUTION OF INCOME
By JOHN CREEDY*
I. THE AGE DISTRIBUTION AND AGGREGATION
One of the many 'stylized facts' which has aroused some interest in the past is
the relative stability of the personal distribution of incomealthough this has
always received less attention than the functional distribution.' Some early
evidence for this stability was presented in the well known works of Pareto [19]
and Gibrat [12],2 who nevertheless had fundamentally different explanations for
the genesis of the income distribution.3
More recently Thatcher [22] has collected information about the quantiles and
first and ninth deciles expressed as percentages of the median from 1886 to 1968
for both manual males and females. These are roughly constant throughout the
period.The stability of the personal distribution is in fact an implicit assumption of
many macroeconomic models of economic behaviour. It is nevertheless important
to stress that although the parameters of the distribution (or some measure of
inequality which is independent of the units of measurement) may not change from
one year to the next there is a great deal of mobility within the quantiles of the
distributions. The simple observation that not all the changes will have been
anticipated, so that 'adjustment' may not be complete (or the new level may not
be regarded as stable) is, of course, basic to much of the postwar work on aggre-
gate consumption.
This mobility also has important implications for normative evaluations of the
degree of 'inequality' which exists since, for example, the length of time during
which people may remain in the lower deciles of the distribution is a crucial factor
in considering lifetime equity.4 This will also influence the ability to accumulate
wealth, of course.
The main purpose of the present paper, however, is to examine the implications
for the analysis of the complete distribution of systematic changes in the income
distribution over age. The fact that changes in the distribution over the life-cycle
in a large variety of professions display the same basic characteristics offers the
possibility of obtaining some quite useful general results for the aggregate distribu-
* I am very grateful to Alan Brown for help and encouragement in the preparation of this
paper.It is rather ironic that in the first (1933) volume of Econometrica Arthur Bowley [3]
attempted to direct attention to the study of distributions in economics (of prices, incomes, etc.)
whereas much of the research since then has been devoted to the analysis of the so-called
Bowley's law' of the stability of factor shares.
2 After reviewing the evidence (mainly continental work) Gibrat concluded ([12] chapter
Il, p- 190) 'Dans l'état actuel de la technique statistique il est impossible d'admettre que
l'inégalité des fortunes ou celles des revenus présentent une tendance vers une augmentation
ou une diminution'.
An interesting discussion of Pareto's ideas is contained in Brown [4], and a further
examination of his work in Creedy [8].
We may in some circumstances regard an increase in the dispersion at any time to be
desirable. See Hart [13] and Creedy [8].
91
To continue reading
Request your trial