2374 intergenerational mobility: methods of analysis

The criteria used to aggregate occupations
into social classes are in many cases more prag
matic than theoretical, but classifications can be
evaluated in terms of both homogeneity and
structure. Homogeneity refers to the extent that,
within an aggregate of occupations (social class),
there are no barriers between any origin occupa
tion and any destination occupation. Structure
refers to pattern and strength of the barriers
between origin and destination classes. These
barriers should be similar for the occupations
constituting the classes. When disaggregated
information is available, the assumption that
class boundaries (barriers) are in agreement with
the requirements of homogeneity and structure
is testable.
The following list makes use of the seven
class version of the Goldthorpe class schema:

1 Classes I II: professionals, administrators
and officials, managers.
2 Class III: routine non manual employees.
3 Class IVab: small proprietors and artisans.
4 Class IVc: farmers.
5 Class V VI: technicians, supervisors, and
skilled manual workers.
6 Class VIIa: semi and unskilled workers not
in agriculture.
7 Class VIIb: semi and unskilled workers in

Table 1 presents an intergenerational mobi
lity table based on data used by Breen and
Luijkx (2004). The table is an average table
based on 89 mobility tables for men from
France, Germany, Great Britain, Hungary,
Ireland, the Netherlands, Poland, and Sweden
between 1970 and 2000. The number of cases in
the table is rescaled for reasons of exposition
from more than 250,000 to 1,000.
We distinguish origin O with subscript i
and destination D with subscript j. Let
ffijgi 1; :::; I; j 1; :::; J be the observed
frequencies for each cell i; j in the I J
mobility table of origin by destination (in this
case I J 7); fij is the observed number of
people with origin i and destination j. The
observed row totals are designated by fi, the
observed column totals by fj, and the grand
total by f or N.


Using the data in Table 1, we can compute
measures of absolute mobility. We observe that
340 out of 1,000 men are in the same class as
their fathers. In other words, 34.0 percent of
the men are immobile and thus the mobility
rate in the table is 66.0 percent. The mobility
rate (or gross change) depends on the number
of classes being distinguished, as can be seen
from Table 2. In Table 2, we combine the
seven classes into three: (1) I II, (2) III, IVab,
IVc, and V VI, and (3) VIIa and VIIb. Based
on these data, we now conclude that 50.8 per
cent (9.5 33.3 8.0) of the people are
immobile and that the mobility rate is 49.2
percent. Within the mobile, we can distinguish
those who are upwardly mobile, going from
(3) to (1) or (2) and from (2) to (1) in total
31.2 percent; and those who are downwardly
mobile, going from (1) to (2) or (3) and from

Table 1 A mobility table

O(rigin) I II III IVab IVc V VI VIIa VIIb Total
I II 95 14 9 1 24 11 1 155
III 26 9 4 1 16 8 1 65
IVab 28 9 18 1 21 13 1 91
IVc 25 9 10 55 40 40 11 190
V VI 67 21 16 1 102 44 2 253
VIIa 38 16 11 2 70 53 3 193
VIIb 5 3 2 2 17 16 8 53
Total 284 81 70 63 290 185 27 1,000