Using Pseudo Cohorts to Track Changes in the Qualifications of

R ESEARCH
Using Pseudo Cohorts to Track Changes in
the Qualifications of National Populations
Steven McIntosh
Centre for Economic Performance
London School of Economics
Research Report RR621
Research Report
No 621
Using Pseudo Cohorts to Track Changes in
the Qualifications of National Populations
Steven McIntosh
Centre for Economic Performance
London School of Economics
The views expressed in this report are the authors’ and do not necessarily reflect those of the Department
for Education and Skills.
© Centre for Economic Performance 2005
ISBN 1 84478 419 3
1
Executive Summary
Key Points
•
The results reveal improvements over time (between 1985 and 2003)
in the level of qualifications held by young people in the UK, both within
cohorts as young people get older, and also between cohorts as new
groups of better qualified individuals join the labour force.
•
In terms of reducing the number of young people below level 2, starting
from a higher base, each cohort observed in the UK has reduced this
number to a greater extent than in France or Germany.
•
Within cohorts, however, as young people age from their early twenties
to their early thirties, the amount of movement out of the below level 2
group is similar in the UK to what is in France, and both countries trail
Germany.
•
Higher up the skills distribution, at level 3 and above, the improvement
across cohorts in less than in France. Germany shows no cross-cohort
improvement, although this is due to all observed cohorts in Germany
achieving a high proportion at level 3 and above.
•
In terms of upgrading within cohorts as young people age from their
early twenties to their early thirties, there is more movement up to level
3 and beyond in both France and Germany, compared to the UK.
•
In the former two countries, the proportion at level 3 via the vocational
route increases markedly as each cohort ages, in every cohort. In the
UK, however, the proportion at level 3 via the vocational route
increases by at best two percentage points as the various cohorts age.
This seems to be a key reason for the greater upgrading of skills within
cohorts in France and Germany compared to the UK.
•
Even if the UK continues to make improvements between cohorts, with
each new cohort entering the workforce becoming successively better
qualified, if the existing workforce is not being upgraded, it will take
many years before the new cohorts work through the system and fully
replace the existing lower qualified cohorts in the workforce. In other
words, without further improvements in upgrading the skills of the
existing adult workforce, the UK is unlikely to catch up with France and
Germany.
2
Detailed Findings
Pseudo cohorts are artificially created data sets constructed from repeated
cross-sections. For example, the first cohort studied here are aged 19-21 in
1985, 25-27 in 1991 and 31-33 in 1997. Individuals of these ages found in
cross-section data sets in the relevant years are therefore extracted to form
the pseudo cohorts. Although the actual individuals surveyed will differ at
each point in time, they will be representative of the full cohort in the
population.
The use of pseudo cohorts allows us to analyse to what extent the growth in
qualifications held by the population is caused by skill upgrading as
individuals age, and to what extent it is caused by new, more highly educated
cohorts entering the workforce to replace older, less well-qualified workers.
The analysis was undertaken for the UK, France, Germany and the US.
Three cohorts are considered. Cohort 1 is aged 19-21 in 1985, and is then
observed again in 1991 when its members are aged 25-27, and in 1997, when
they are aged 31-33. Cohort 2 is constructed so that its members are aged
19-21 in 1991, 25-27 in 1997 and 31-33 in 2003. Members of the third and
final cohort created are aged 19-21 in 1997, and 25-27 in 2003. The results
for the UK will be considered first, and then the results in the other countries
will be compared.
UK
Cohort 1 reveals large numbers of low qualified (below level 2) workers at the
age of 19-21. However, there is some upskilling as the cohort members age,
with the number of low-qualified individuals in the cohort falling by about a
third by the time they are aged 31-33.
The largest growth in qualifications as the cohort members age is observed
above level 3, with the proportions at levels 2 and 3 remaining roughly
constant. The most plausible story behind these results is that there is a
3
‘bumping up’ of levels, with those initially below level 2 moving up to level 2,
while at the same time individuals move out of level 2 into level 3, where
again there are individuals moving out into the higher qualification levels.
There is some evidence that it is individuals with academic, rather than
vocational, qualifications at levels 2 and 3 who are more likely to move up the
qualifications hierarchy as they age.
The main difference between men and women is that the proportion of lowqualified women does not fall as quickly as that of men, as the cohort
members get older.
The key reason for this seems to be fewer women
acquiring intermediate vocational qualifications than men.
Turning to cohort 2, the proportion failing to reach level 2 by the age of 19-21
is significantly lower than in the earlier cohort. Not only is there significant
qualifications upgrading within a particular cohort, therefore, but there is also
significant qualifications upgrading between cohorts, with younger cohorts
being better qualified.
The pattern of upgrading within Cohort 2 is similar to that observed in the
earlier cohort, thus again revealing quite extensive upgrading of qualifications
amongst the cohort members. However, the rate of decrease of the stock of
low-qualified individuals across time periods does seem to be slower in the
second cohort than in the first, presumably because a group of hard-toupgrade individuals is increasingly reached as the size of the low-qualified
stock itself becomes smaller.
The other main difference compared to cohort 1 is that the proportion with
level 3 qualifications falls as the members of the second cohort age. The
reason for this is a higher proportion of individuals reaching level 3 by the age
of 19-21 in Cohort 2, principally via the academic route, who then go on to
acquire higher qualifications as they get older.
Cohort 3 is only observed twice, but the pattern within the cohort is similar for
4
these two observations as was found for the first two cohorts. The main
differences are a further slowing of the movement out of the ‘below level 2’
group as the cohort members age, but increased upgrading amongst those
who reached levels 2 and 3 by the age of 19-21, particularly amongst those
who followed the academic route. The outcome of this upgrading is very large
increases in the proportion reaching level 4 or above as the cohort members
age.
Comparing Cohort 3 to the earlier cohorts reveals further upgrading across
cohorts, with the size of the low-qualified group, for example, being smaller
again than observed in cohorts 1 and 2.
France
Comparing Cohort 1 in France with their UK counterparts, there is a lower
stock of low-qualified (below level 2) individuals at the age of 19-21 in France
compared to the UK.
Despite this, the French still achieve faster upgrading at all qualification levels
(i.e. a larger decrease in the size of the ‘below level 2’ group and a larger
increase in the size of the ‘level 3 or above’ group) as the cohort members
progress into their mid-twenties, thus further increasing the gap between the
qualifications levels of Cohort 1 in the UK and France.
The gap does then close somewhat as the cohort members age further, into
their early thirties, since the UK cohort members continue to upgrade,
whereas there is much less movement in the French qualification levels
between these ages.
Comparing Cohort 2 with Cohort 1, there is a greater improvement across
cohorts in terms of the proportion failing to reach level 2 by the age of 19-21,
in the UK compared to France. However, at higher levels, there is more
improvement between the 2 cohorts in France than in the UK.
5
France also continues to have a faster upgrading than the UK of individuals’
qualifications within Cohort 2, particularly for men, as the cohort members
age, from their early to their mid-twenties, again at all qualification levels.
Similarly, amongst those in Cohort 3, there is again more upgrading of
qualifications, at all levels, in France than in the UK.
Comparing between Cohorts 2 and 3, as between Cohorts 1 and 2, there is
more improvement between cohorts in terms of reducing the numbers failing
to reach level 2 in the UK than in France, but higher up the distribution,
France sees greater improvement in the numbers at level 3+ across cohorts
than the UK.
Germany
There are no data available for Cohort 1 in Germany. Considering Cohort 2,
Germans of this age are already significantly better qualified than their UK
counterparts by the age of 19-21.
Nevertheless, there is still more
qualifications upgrading within this cohort as the members get older in
Germany than in the UK, at all points of the qualifications distribution. The
initial gap between the qualifications of Cohort 2 in the two countries at age
19-21 therefore simply widens as the group get older.
Similarly within Cohort 3, there is significantly more qualifications upgrading
as the cohort members age in Germany than in the UK.
However, the UK does make significant improvements across cohorts,
comparing the stock of qualifications of Cohort 3 to Cohort 2 when both were
aged 19-21. In Germany, Cohort 3 are no better qualified than Cohort 2 at the
same point in the lives.
USA
Only consistent data for Cohort 3 are available, and so no between cohort
6
effects can be calculated for the US.
Looking just at Cohort 3, therefore, the existing qualification profile is quite
similar to that of the UK.
Within Cohort 3, there is more upgrading of qualifications at the lower end (out
of the ‘below level 2’ group) in the US than the UK, but more upgrading of
qualifications at the higher end (into level 3 or higher) in the UK.
Summary
The way in which the UK is closing the skills gap on France and Germany is
primarily through improvements in the qualifications levels of successive
cohorts entering the labour market. Thus the UK is seeing greater reductions
in the proportion failing to reach level 2 across cohorts than both France and
Germany, and also greater increases in the proportion reaching level 3 or
above across cohorts than Germany, though not France.
Looking within
cohorts, however, the extent of skills upgrading of young people in their
twenties is greater in both Germany and France in terms of getting people up
to level 3 and beyond, and in Germany only in terms of getting people up to at
least level 2. The vocational route seems to be one way that these countries
achieve greater upskilling within cohorts as young people age from their
teenage years to their thirties.
7
Introduction
Recent work undertaken for the update to the Skills Audit has estimated the
proportions of national populations to be at each of the levels on the UK
national qualifications framework.
This has been done for a number of
countries, namely the UK, France, Germanty, the US and Singapore, using
the most recent data available in each country, usually 2002 or 2003. The
results from this update could then be compared with those of the original
Skills Audit and the first update, undertaken using data from 1994 and 1998
respectively.
Such comparisons provide information on the changes over
time, and thus allow us to see how the stock of qualifications is changing in
each country. One question that cannot be answered using such repeated
cross-sections of data covering the whole population, however, is whether any
changes observed over time, for example a rise in the number of
qualifications held, are due to increased numbers of qualifications being held
by individuals who were, and remain, in the workforce, or whether they are
due to older, less qualified, workers leaving the workforce, to be replaced by
younger, more qualified workers. Another way of putting this question is to
ask whether the increase in qualifications attained is occurring within
particular cohorts of workers, or whether it is occurring between cohorts. Of
course, the likely answer is that both effects are occurring, but it would be
interesting to find out the relative size of each.
The answer will provide
important information regarding the supply of skills to each country. If most of
the increase in qualifications attainment is occurring between cohorts, then
achieving an increase in the supply of skills will be a matter of waiting until the
older cohorts are replaced by the younger, better qualified cohorts. If the
increase in qualifications attainment is occurring within cohorts, however,
such that each cohort of workers is acquiring qualifications as they age, then
the supply of skills could be expected to be more responsive to any increase
in the demand, since individuals of working age would be raising their
qualification attainment, and hence their skills.
In order to investigate such questions, longitudinal data that follow individuals
8
over a long period of time would be the most appropriate to use. Analysis of
such data would reveal whether individuals, or cohorts of individuals, increase
their qualification attainment over time as they age, or whether their
qualifications remain reasonably constant over their working lives, so that the
supply of skills in the economy is only increased when this cohort retires from
the workforce to be replaced by a younger, more qualified cohort of
individuals. Unfortunately, no such data set is available in most countries.
The analysis in this report will therefore proceed by creating pseudo cohorts
of individuals using national data sets from each of the countries of interest. A
pseudo cohort is created from repeated cross-sectional data sets, by taking
the data for a particular age group from each data set. For example, the first
pseudo cohort to be studied below is aged 19-21 in 1985. Thus all individuals
who fall within this age range in the 1985 survey are put into the pseudo
cohort. In 1991, 6 years later, this cohort will be aged 25-27, and so all
individuals who are in this age range in the 1991 survey are added to the
pseudo cohort as the second observations in the time dimension. Of course,
since the data sets used are repeated cross-sections rather than panel data
sets, the actual 25-27 year olds interviewed in 1991 will not be the same
people as the 19-21 year olds surveyed in 1985. However, if the data sets
used are nationally representative, as all data sets used here are, then the
pseudo cohort created will be representative of the true cohort of this age in
the population as a whole, and estimates of the proportions at each
qualification level using data from these pseudo cohorts will be unbiased
estimates of the true proportions in the population of this age.
In the analysis that follows, 3 pseudo cohorts are created. The first two are
observed at three points in time, while the last one is observed just twice. The
use of the three cohorts allows us to see whether each is becoming more
qualified as the individuals in it age, and also, comparing across cohorts at the
same points in their lives, whether the younger cohorts are more qualified
than the older cohorts. Cohort 1 is aged 19-21 in 1985, and is then observed
again in 1991 when its members are aged 25-27, and in 1997, when they are
aged 31-33. Cohort 2 is constructed so that its members are aged 19-21 in
1991, 25-27 in 1997 and 31-33 in 2003. Members of the third and final cohort
9
created are aged 19-21 in 1997, and 25-27 in 2003.
UK
Cohort 1
Figure 1 shows the results for the first cohort, aged 19-21 in 1985, for males.
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1985 (age19-21)
1991 (age25-27)
1997 (age 31-33)
vo
c
l
e
-o
ve
fw l3
hi
ch
ac
vo
>l c
- o ev e
fw
l
hi 3
ch
ac
vo
c
<l
ev
el
2
l
- o eve
fw l2
hi
ch
ac
proportion
Figure 1: Cohort 1 - Males (UK)
The chart clusters the three observations in time, 1985, 1991 and 1997,
together at each qualification level. The levels studied are “below level 2”,
“level 2”, “level 3” and “above level 3”. The chart shows the results when the
last three categories are split into academic and vocational qualifications at
these levels, as well as the cumulative total at each level.
The first thing to note, at the left-hand side of Figure 1, is the declining
proportion of this cohort of individuals failing to reach level 2 on the
qualifications framework. This reveals that the members of this cohort are
acquiring more qualifications, at level 2 or above, as they age. The changes
are large, with 64%1 of the cohort failing to reach level 2 in 1985, compared to
51% in 1991 and 41% in 1997. The increase in attainment is therefore fastest
when the cohort are still aged in their early and mid-twenties. However, even
when they begin to move into their thirties, in 1997, the decline in the
1
The actual proportions are given for all figures in this report in the tables in Appendix B.
10
proportion failing to reach level 2 is still significant.
Looking across the figure, it is clear that the largest increase in attainment as
the cohort age is at the highest level, above level 3. It is unreasonable to
expect that individuals are jumping straight from the “below level 2”category to
the “above level 3” category (though of course we cannot prove that this is not
happening, since we do not actually observe the same people at the different
points in time, and so do not know what the qualification level was in 1985 of
an individual observed at above level 3 in 1997). It seems more plausible
that, as the cohort age, some move out of the “below level 2” category into the
“level 2” category, some move out of the “level 2” category into the “level 3”
category and some progress from the “level 3” category to the “above level 3”
category. The flows into and out of levels 2 and 3 seem to broadly cancel
each other out, as the proportions at these levels do not change much over
time. When the middle levels are split into their academic and vocational
components, however, an interesting, though not surprising, difference
emerges.
Although the changes are not large, Figure 1 shows that the
proportion at level 2 through an academic qualification in this cohort is falling
over time, while the proportion at level 2 through a vocational qualification is
rising over time. This suggests that, although the flows into and out of level 2
qualifications in this cohort are broadly cancelling themselves out in
aggregate, when split into type, the inflows into level 2 from below seem to be
more likely to be acquiring vocational qualifications, while the outflow from
level 2 to higher levels is more likely to be amongst individuals who held
academic qualifications at the former level.
The outflow of academically-
qualified individuals, and the inflow of vocationally-qualified individuals, mean
that the balance between academic and vocational qualifications at levels 2
and 3 changes over time as the cohort ages, with vocational qualifications
being increasingly represented, to such an extent that by 1997, more people
at level 3 hold vocational qualifications than academic qualifications.
As mentioned above, however, the changes over time just noticed at levels 2
and 3 are not large, compared to the “above level 3” category. It is therefore
at the highest level that attainment amongst males in this cohort is increasing
11
fastest. In 1985, when they are aged 19-21, just 3% are qualified above level
3. By 1991, when they are aged 25-27, this proportion has risen to 18%, and
it is still rising quite steeply as they move into their thirties, reaching 26% by
1997. Amongst these, academic qualifications dominate, although vocational
qualifications have played their part in rising attainment at this level. Hence,
in 1997, 17% of this cohort of men have reached above level 3 via the
academic route, and a further 9% have reached this level via the vocational
route.
Figure 2 shows the same analysis for females in this age cohort.
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1985 (age19-21)
1991 (age25-27)
voc
- of which ac
>level 3
voc
- of which ac
level 3
voc
- of which ac
level 2
1997 (age 31-33)
<level 2
proportion
Figure 2: Cohort 1 - Females (UK)
The pattern of results is very similar for females as it was for males. When
the cohort are first observed, in 1985, 64% are qualified at below level 2,
exactly the same percentage as amongst the men. This percentage also falls
for the women in this cohort, although not quite as far as it did for the men,
reaching 53% by 1991 when they are aged 25-27, and 46% by 1997, when
they are aged 31-33. Again as for the men, the category that benefits most in
terms of increased numbers is the “above level 3” category, although it again
seems reasonable to assume that there has been an upgrading of
qualifications across all levels, rather than women in this cohort jumping
straight from the “below level 2” category in 1985 to the “above level 3”
category in 1997.
12
One comment that can be made when comparing males and females is that
the proportion qualified at below level 2 does not fall quite as fast for women
as for men. Despite a similar starting position of 64% at this level for both
genders when aged 19-21 in 1995, the percentage falls to 46% by 1997 when
the cohort are aged 31-33 for women, compared to the 41% figure noted
above for males. Although the difference is not large enough to devote too
much attention to, a possible explanation could be that women appear to be
less likely to achieve vocational qualifications after leaving full-time education.
Thus, although the figures for women reaching levels 2 and 3 via the
academic route compare favourably to those of men (15% of the female
cohort have reached level 2 via the academic route by 1997, compared to
11% of men, with the figures at level 3 being 7% and 6% respectively), when
we look at the vocational route for reaching levels 2 and 3, the numbers show
4% of the female cohort at level 2 by 1997, compared to 8% of the male
cohort, and similarly 4% of the female cohort at level 3 by 1997, compared to
7% of the male cohort. It should also be pointed out that part of the female
cohort’s relative success in the academic route at levels 2 and 3 is due to
fewer numbers progressing to the highest levels (level 4 and above) via the
academic route (14% of the female cohort and 17% of the male cohort),
although women perform as well as males in terms of high level vocational
qualifications, helped by the large number of women acquiring a nursing
qualification.
Figure 3 below shows the results of a similar analysis when males and
females are pooled into the same data set. Since the results are quite similar
for men and women, the pooled results do not look too different to the results
for either gender on their own, and therefore the pooled results will not be
commented up further, in each case in this report.
13
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1985 (age19-21)
1991 (age25-27)
voc
- of which ac
>level 3
voc
- of which ac
level 3
voc
- of which ac
level 2
1997 (age 31-33)
<level 2
proportion
Figure 3: Cohort 1 - All (UK)
Cohort 2
Figure 4 shows the qualification achievements for males in the second cohort,
initially aged 19-21 in 1991, and observed further in 1997 and 2003. The
pattern of results is quite similar to that in Figure 1 for the older cohort,
although there are some important differences, particularly at level 3.
Starting again at the lowest level, below level 2, the percentage of the new
cohort at this level at the age of 19-21 in 1991 is 50%, which is considerably
lower than the equivalent percentage in the older cohort at the same age,
which was 64%.
The analysis above therefore revealed that there was
significant upgrading of qualifications within the older cohort over time, while
this new result shows that there is also considerable upgrading across
cohorts, with a much smaller proportion of the younger cohort failing to reach
level 2.
14
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1991 (age 19-21)
1997 (age 25-27)
2003 (age 31-33)
<l
ev
el
2
l
- o eve
fw l2
hi
ch
ac
vo
c
l
e
-o
ve
fw l3
hi
ch
ac
v
>l oc
- o ev e
fw
l
hi 3
ch
ac
vo
c
proportion
Figure 4: Cohort 2 - Males (UK)
Of course, there is also upgrading within the new, younger cohort as well. As
they age, the proportion of men in this cohort still failing to reach level 2 falls
to 40% in 1997 when they are aged 25-27, and to 38% in 2003, when they are
aged 31-33. This rate of decrease in the numbers at this level (i.e. the rate of
increase in qualifications upgrading) is slower than in the older cohort, but this
is to be expected since the low qualified group is smaller in size in Cohort 2,
and so we would expect to reach the lower ability individuals more quickly.
Nevertheless, the rate of increase in qualifications does seem significantly
lower in the second period, between the ages of 25-27 and 31-33, in the
second cohort (with a 2.8 percentage point fall in the number failing to reach
level 2), compared to the first cohort (a 9.5 percentage point fall). It would
therefore seem that, as the number failing to reach level 2 by the age of 19-21
falls, it is going to become more difficult to make inroads into reducing this
number further as they age.
At level 2, the situation amongst the younger cohort, Cohort 2, is almost
exactly the same as in the older cohort, with the percentage of men at this
level remaining fairly constant at around 20% at all 3 points in time. Again, it
seems reasonable to suppose that this is due to the inflows to this group from
15
the low-qualified group almost exactly balancing the outflows to higher
qualification levels. Again, evidence consistent with this assertion is that the
proportion at level 2 via the academic route falls as the cohort age, with the
academically able being more likely to progress to higher levels, while the
proportion at level 2 via the vocational route increases as the cohort age, with
the initially low-qualified being more likely to acquire vocational qualifications
to move them up the skills ladder.
As mentioned above, the key differences between Cohorts 1 and 2 occur at
level 3. Whereas in the first male cohort the proportion at level 3 was quite
constant at each point in time, falling less than three percentage points as
they aged, in the second, younger, cohort, the fall in the proportion at level 3
as the cohort age is much more dramatic, falling by more than half from 27%
to 13%. The reason for the difference is the much larger numbers having
level 3 qualifications in the first place by the age of 19-21 in the younger
cohort. The larger numbers at this level at this age compared to the earlier
cohort consists quite evenly of extra numbers holding both academic and
vocational qualifications, both of which are about 5 percentage points higher
than in the first cohort. The obvious reason why the numbers at level 3 then
decline as the cohort members age is that the inflows from men moving up,
from level 2 or below, to level 3 are not matching the outflows of men from
level 3 into the highest qualification levels. This is revealed by the large
numbers reaching the higher levels, 31% of the male cohort by 2003, which
exceeds the 26% of the first, older, cohort.
As expected, most of this
progression is via the academic route, and first degrees in particular. The
proportion of the second male cohort reaching the highest levels by their
earlier thirties via the vocational route is actually slightly lower than in the first
cohort. The final point to note about the progression the cohort make from
level 3 to the higher levels is that it largely takes place before the age of 25,
as we would expect if most individuals making this progression are following a
straight academic route through to degree level.
Turning to the women in cohort 2, the situation for them is displayed in Figure
5. The story is very similar to that of males. Again, therefore, the second
16
cohort has a much lower number failing to reach level 2 by the age of 19-21
(50% compared to 64%), and this number then declines further as the cohort
age, although less quickly than in the first cohort when there were more
women to make the progression.
Interestingly, the proportion of women
upgrading from below level 2 in the second cohort is very similar to their male
counterparts in the second cohort, unlike in the first cohort when slightly more
men made the progression. The outcome of this shows up in the level 2
results, where the proportion of women reaching this level via the vocational
route rises as the cohort get older to a greater extent than in the first cohort.
As usual, this rise in vocational level 2 qualifications is counter-balanced by a
fall in the proportions having academic level 2 qualifications as their highest,
leaving the overall proportion at level 2 reasonably constant over time.
0.6
0.5
0.4
1991 (age 19-21)
1997 (age 25-27)
0.3
0.2
0.1
0.0
2003 (age 31-33)
-o
<l
ev
el
lev 2
f w el 2
hi
ch
ac
vo
c
l
e
-o
ve
fw l3
hi
ch
ac
v
>l oc
- o ev
f w el 3
hi
ch
ac
vo
c
proportion
Figure 5: Cohort 2 - Females (UK)
At level 3, the situation for women in the second cohort mirrors that of men,
with a steep fall in the proportion having level 3 as their highest qualification
as they get older, which was not observed in the first cohort, principally
because more women reach level 3 by the age of 19-21 in the fist place in the
second cohort.
As was the case for men, most of this decline in the
proportion at best at level 3 is amongst women qualified at this level via the
academic route, and then going on to acquire qualifications at a higher level.
17
This results in larger numbers of women above level 3 in the second,
compared to the first, cohort (30% by the age of 31-33, compared to 24% in
the first cohort), most of which is achieved by the age of 25-27 when the
cohort are observed for the second time. Again as for men, this improvement
in higher level qualifications relative to the older cohort is achieved exclusively
via the academic route, the proportion of the second female cohort having
vocational qualifications at the highest level actually being lower in the second
than in the first cohort.
The pooled results for men and women combined are shown in Figure 6.
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1991 (age 19-21)
1997 (age 25-27)
2003 (age 31-33)
-o
<le
v
el
2
lev
f w el 2
h ic
ha
c
vo
c
l
- o eve
fw l3
h ic
ha
c
vo
>le c
v
-o
f w el 3
h ic
ha
c
vo
c
proportion
Figure 6: Cohort 2 - All (UK)
Cohort 3
The final cohort to be followed are aged 19-21 in 1997. They are therefore
only observed once more, in 2003 aged 25-27. The other difference with the
analysis of this cohort compared to the earlier ones is that, since they are only
observed after 1996, more information on their qualifications is available
following the change in the LFS questionnaire in 1996, and so the calculations
for the third cohort are performed using the new, 1996, classification (full
details are in Appendix A).
Figure 7 displays the results for the male members of the third cohort.
18
Figure 7: Cohort 3 - Males (UK)
proportion
0.4
0.3
1997 (age 19-21)
0.2
2003 (age 25-27)
0.1
vo
>l c
- o ev
f w el 3
hi
ch
ac
vo
c
vo
c
l
- o eve
fw l3
hi
ch
ac
<l
ev
el
2
l
- o eve
fw l2
hi
ch
ac
0.0
The figure reveals that as we move to this still younger cohort, further
improvements have been made in terms of men failing to reach level 2 by the
age of 19-21. This percentage is now down to just 33% in the third cohort.
There is still some progression within the cohort, so that the percentage has
fallen further to 26% by the time the cohort are aged 25-27, this change over
time being slightly lower than in the previous cohorts, as a harder core of lowachievers is increasingly reached.
A key difference between the third cohort and the earlier two cohorts occurs at
level 2. Whereas in the older cohorts the proportion achieving at best level 2
is fairly stable over time, with inflows to this level matching outflows, this is not
the case in the third cohort, where a steep fall in the proportion at this level is
observed, from 29% when aged 19-21 in 1997 to 19% when aged 25-27 in
2003. The explanation seems to lie in the higher proportion at level 2 by the
age of 19-21 in the first place in the third cohort. It is encouraging that the
greater numbers reaching level 2 by this young age are still able to progress
to higher levels as they get older, however, and this large outflow to higher
levels greatly exceeds the now smaller inflow of individuals progressing from
below level 2, which thus explains the declining proportion at level 2 as the
cohort age. This outflow from level 2 is largely made up of individuals who
reached level 2 via the academic route, the academic level 2 proportion
19
declining between 1997 and 2003 much more quickly than the vocational level
2 proportion.
Progression continues across the board in level 3. Despite the larger number
leaving level 2 as just described, and so a larger inflow into level 3 than in
previous cohorts, the proportion observed with a level 3 qualification as their
highest still declines steeply between 1997 and 2003, from 33% to 19%. This
reveals that the outflow into high level qualifications from level 3 still remains
much larger than the increased inflow from lower levels, with the outflow again
dominated by those with academic qualifications. However, boosted by this
larger inflow from level 2, and despite the large outflows to high level
qualifications, there is still a higher proportion of the third cohort reaching level
3 as their highest qualification, compared to the earlier cohorts, by the age of
25-27 (19% in cohort 3 compared to 15% and 14% respectively in cohorts 1
and 2).
Also, perhaps encouragingly, this is increasingly being achieved
through vocational qualifications, with 12% of the third cohort reaching level 3
at best via the vocational route by the age of 25-27, compared to 8% in each
of the first two cohorts.
We observe a greater proportion of the third cohort at the highest levels than
in the previous cohorts, as expected given the high outflows from the level 3
academic group noted above. Thus, by the age of 25-27, 37% of the third
male cohort are qualified at level 4 or above, compared to 26% in cohort 2
and just 18% in cohort 1. Given that the age difference between each cohort
is just 6 years, these are very large increases.
The rise in high level
qualifications across cohorts continues to be in academic qualifications, the
proportion of the third cohort having above level 3 vocational qualifications
again declining relative to the previous cohort at the same stage in their
working lives.
Figure 8 displays the results for the females in Cohort 3. It reveals that the
proportion of the third cohort who fail to reach level 2 by the age of 19-21 has
again fallen, and is now down to just 30%. However, of some cause for
concern is that this proportion hardly changes as the cohort age, being still
20
26% when they are aged 25-27 in 2003. Thus, although the group of lowachieving school leavers is declining with successive cohorts who leave fulltime education, those who remain are proving more difficult to reach in terms
of post-school education or training.
Figure 8: Cohort 3 - Females (UK)
proportion
0.4
0.3
1997 (age 19-21)
0.2
2003 (age 25-27)
0.1
-o
<l
ev
el
lev 2
f w el
hi 2
ch
ac
vo
c
l
e
-o
ve
fw l3
hi
ch
ac
v
>l oc
- o ev
f w el
hi 3
ch
ac
vo
c
0.0
As for the men, for the first time amongst the three cohorts, Cohort 3 display
significant changes at level 2 over time as they age. The fact that many more
than in earlier cohorts are reaching level 2 in their teen years, means that
there are fewer to inflow into this group in their twenties, while there are more
individuals to outflow to higher levels. Thus, the proportion with a level 2
qualification as their highest falls from 32% to 21% amongst the women in
Cohort 3. Most of this outflow is the academically qualified moving up to
higher levels.
At level 3, we again observe higher numbers than in earlier cohorts reaching
this level by the age of 19-21, up to 33% in cohort 3.
This group then
continue their progression to higher levels by the age of 25-27, so that the
large outflow from level 3 to higher levels again exceeds the numbers moving
up from level 2, leaving a decreased proportion at best at level 3 by the age of
25-27 in 2003. Nevertheless, this proportion, 17%, remains higher than in the
21
earlier cohorts, and again as for the men, this is primarily due to increased
numbers reaching level 3 via the vocational route in the youngest cohort.
Thus, in cohort 3, 10% of the females have a vocational level 3 qualification
as their highest qualification by the age of 25-27, compared to 5% and 6% in
the two earlier cohorts at the same age.
Finally, at the highest level, 36% of the third cohort of women are qualified at
level 4 or above by the age of 25-27 in 2003. This again represents a huge
increase on the earlier two cohorts, who had equivalent percentages of 17%
and 25% at the same age. In a final similarity to the men, this increase is
being driven entirely by academic qualification acquisition, with the proportion
being qualified at level 4 or above via the vocational route actually being lower
in cohort 3 (6%) than in cohorts 1 or 2 (both 7%).
Finally for the UK, the results for men and women combined in Cohort 3 are
shown in Figure 9.
Figure 9: Cohort 3 - All (UK)
0.3
0.2
2003 (age 25-27)
el 2
l
- o evel
fw
2
h ic
ha
c
vo
c
l
- o evel
fw
3
h ic
ha
c
vo
>le c
v
-o
f w el 3
h ic
ha
c
vo
c
0.1
0.0
1997 (age 19-21)
<le
v
proportion
0.4
22
France
Cohort 1
We turn now to the other countries in the study. The main interest is how they
compare to the UK, and it is such points of comparison that will be discussed
in what follows, rather than a full description for each country as was given for
the UK above.
Figure 10 below begins an identical analysis for France, considering first the
cohort of men who were aged 19-21 in 1985.
Figure 10: Cohort 1 - Males (France)
0.6
proportion
0.5
0.4
1985 (age19-21)
0.3
0.2
1991 (age25-27)
0.1
1997 (age 31-33)
vo
>l c
- o ev
f w el 3
hi
ch
ac
vo
c
vo
c
l
- o eve
fw l3
hi
ch
ac
-o
<l
ev
el
2
lev
f w el 2
hi
ch
ac
0.0
The first point to note, on the left hand side of the figure, is the lower
proportion of 19-21 year French men in 1985 who had failed to reach level 2
by this age (40%) compared to their UK counterparts (64%). Despite this
smaller stock of low qualified young men in France, there is still a further,
larger, fall in the proportion below level 2 amongst this cohort by the age of
25-27 in 1991 (a 15 percentage point fall to 25%), compared to the UK, where
there was a 13 percentage point fall to 51%. France thus gets more of her
initial low-achievers up to at least level 2 when aged in their twenties, despite
the target group being smaller and therefore more likely to be from the lower
end of the ability distribution. Therefore France, already ahead in terms of the
23
number at level 2 or above at age 19-21 (in 1985), further increases this lead
amongst 25-27 year olds in 1991. The UK does, however, manage to close
the gap again to a certain extent as the cohort age between 25-27 and 31-33.
Thus, between these ages, there is a further 10 percentage point fall, to 41%,
in the proportion of UK men in this cohort failing to reach level 2, compared to
just a 1 percentage point fall in France. Most of the improvement of the low
school achievers in France therefore occurs by the time they are aged in their
mid-twenties, whereas in the UK it is ongoing into their thirties.
In the UK, as Cohort 1 aged, the proportions at levels 2 and 3 remained
broadly constant, as the inflow from lower qualification levels was matched by
an outflow to higher qualification levels. Amongst French men in Cohort 1,
however, there is a clear fall in the proportion at level 2 between the ages of
19-21 and 25-27, suggesting that the outflows (i.e. progression to higher
levels) is the dominant force at work, despite the high inflow levels of
upgrading from below level 2 noted above.
Since the level 2 group is
dominated by individuals who reached this level via the vocational route
(through obtaining the CAP/BEP), then much of the progression from level 2
to higher levels is by those who reached level 2 via the vocational route,
unlike in the UK, where it is predominately individuals who reached level 2 via
the academic route who go on to progress to higher levels. The last point to
note at level 2 is again that the proportion barely changes between 1991 and
1997 when the cohort age from 25-27 to 31-33, and so most of the upgrading
from level 2 to higher levels takes place when individuals are aged in their
early twenties, similar to the upgrading of those initially below level 2 as
discussed above.
At level 3 in France, there is a growth in the proportion of Cohort 1 qualified at
this level as they age from 19-21 to 25-27, unlike in the UK where this
proportion if anything fell as the cohort aged. The reason for this difference
across countries is the large inflows from level 2 to level 3 in France. The
reason that the level 3 proportion grows in France is NOT therefore because
fewer people progress from level 3 to higher levels in France than in the UK.
In fact, the proportion of this cohort reaching level 4 or above by the age of
24
25-27 is higher in France than in the UK, although it is true that this proportion
continues growing in the UK as the cohort age into their thirties, whereas in
France, as noted for the lower levels above, there is very little qualification
upgrading between the ages of 25-27 and 31-33.
The outcome of the results presented in the last two paragraphs is that there
is more upgrading in France from the ‘level 2 or below’ group into the ‘level 3
or above’ group, as the cohort members age between 19-21 and 31-33.
Between these ages, the proportion qualified at least to level 3 increases by
27 percentage points in France (from 6% to 33%), compared to 20
percentage points in the UK (from 20% to 40%). Thus it was noted above that
France was more successful in upgrading within cohorts at the lower end of
the qualifications distribution (i.e. getting people out of the ‘below level 2’
group), but now we see France is also more successful at upgrading within
Cohort 1 at the higher (level 3 and above) end of the distribution. There is
therefore more upgrading of qualifications going on in France than in the UK
within Cohort 1 as the members age, across all levels of qualifications.
The picture for French women in Cohort 1 is given in Figure 11 below. Much
the same story emerges as for the men. Thus, the proportion of Cohort 1
women failing to reach level 2 by the age of 19-21 in France (37%) is much
smaller than in the UK (64%), and yet despite this, there is still a greater
upgrading to higher levels by the age of 25-27 amongst the French women (a
13 percentage points change, leaving just 24% below level 2) than in the UK
(an 11 percentage point change, leaving 53% below level 2). As with the
men, the UK does partially close the gap as Cohort 1 women reach the age of
31-33, since the proportion below level 2 continues to fall in the UK between
the mid-twenties and the early thirties, whereas it does not fall any further in
France.
25
Figure 11: Cohort 1 - Females (France)
proportion
0.6
0.5
0.4
1985 (age19-21)
0.3
1991 (age25-27)
0.2
0.1
1997 (age 31-33)
-o
<l
ev
el
le 2
f w vel
hi 2
ch
ac
vo
- o lev c
f w el
hi 3
ch
ac
vo
c
- o >lev
f w el
hi 3
ch
ac
vo
c
0.0
Similar effects for French women to their male counterparts are also observed
further up the distribution, with a fall in the proportion with at best level 2
qualifications, a small rise in the proportion with at best level 3 qualifications,
and a big rise in the proportion with high level qualifications, with almost all of
the changes occurring between the ages of 19-21 and 25-27. By the last
point that this cohort are observed (1997 when they are aged 31-33) there is a
higher proportion of French women at level 4 or above (26%) than their UK
counterparts (24%), and thus there has been more upgrading.
Figure 12 shows the position for the pooled sample of men and women in
Cohort 1 in France.
26
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1985 (age19-21)
1991 (age25-27)
1997 (age 31-33)
-o
<l
ev
el
lev 2
f w el
hi 2
ch
ac
vo
- o lev c
f w el
hi 3
ch
ac
v
>l oc
- o ev
f w el
hi 3
ch
ac
vo
c
proportion
Figure 12: Cohort 1 - All (France)
Cohort 2
Figure 13, below, displays the situation for males in the second cohort in
France2. The first thing to note is that the proportion falling below level 2 at
age 19-21 (31%) is still comfortably below the equivalent UK figure for Cohort
2 (50%). However, the gap between the two has closed compared to Cohort
1, where the figures for the proportions at below level 2 amongst 19-21 year
olds were 40% and 64% respectively in France and the UK. Thus, the UK
has knocked 14 percentage points of this low-qualified 19-21 year old rate
between Cohorts 1 and 2, compared to 9 percentage points in France. Of
course, it might have been expected that this would be the case, given that
the larger base of low-qualified individuals in the UK provides more room for
improvement. It is still pleasing that this is the case, however, and the UK is
closing the gap between cohorts in terms of proportions qualified to level 2 or
above.
2
Note that because data could only be obtained until 2001, the last observation on Cohort 2
is actually when they are aged 29-31, rather than 31-33 as in the UK. Given the limited
upgrading after the age of 25-27 observed above in France, this earlier cut-off point in France
should barely affect the results.
27
Figure 13: Cohort 2 - Males (France)
1991 (age 19-21)
1997 (age 25-27)
2003 (age 31-33)
<l
ev
el
2
l
- o eve
fw l2
hi
ch
ac
vo
c
- o leve
fw l3
hi
ch
ac
v
>l oc
e
-o
f w v el 3
hi
ch
ac
vo
c
proportion
0.5
0.4
0.3
0.2
0.1
0.0
As the main Skills Audit report made clear, the point at which the UK lags
France is at level 2. France has a much higher proportion of individuals at
level 2 than the UK, which fully explains the difference in the proportion below
level 2. At higher levels (level 3 or above level 3) the situation in France and
the UK is quite similar. The same is observed in our cohorts here. Thus, in
Cohort 1, similar proportions reach level 3 and above level 3 by the age of 3133. Indeed, these proportions are slightly higher in the UK. By Cohort 2, this
has changed around, however, and the proportions at age 31-33 who are at
best at level 3, and at level 4 or above, are both higher in France than the UK
(16% versus 13%, and 33% versus 31% respectively for levels 3 and 4+).
Thus although the UK displays faster upskilling across cohorts between
Cohorts 1 and 2 in terms of reducing the numbers below level 2, France is
upskilling more quickly in terms of successive cohorts reaching higher
qualification levels. There is therefore not a consistent UK-France difference
in between cohort changes between Cohorts 1 and 2, with the between cohort
changes being larger at the bottom end in the UK, and larger at the top end in
France.
Turning to upskilling within cohorts, it is again true, as it was for Cohort 1, that
there is more qualifications upgrading within Cohort 2 in France than in the
UK. Between the ages of 19-21 and 25-27 in France, the proportion of Cohort
28
2 members qualified below level 2 falls by 13 percentage points (from 31% to
18%), compared to a 10 percentage point fall in the UK (50% to 40%).
Although there is again virtually no upgrading between the ages of 25-27 and
29-31 amongst Cohort 2 in France, there is also less upgrading at this age in
the UK Cohort 2, and so the UK barely closes the gap again on France
through these later age groups.
Figure 13 also makes clear that between the ages of 19-21 and 25-27, there
is not only a large fall in the proportion below level 2, but also in the proportion
exactly at level 2. The subsequent columns make it clear that this was due to
a fall in the proportion with academic level 2 qualifications, who were instead
progressing to higher levels. This should be contrasted with Figure 4 for
Cohort 2 in the UK, where the proportion at level 2 barely changed as the
cohort members aged. This large outflow from level 2 in France in turn raises
the numbers at level 3, and so we do not see the large fall in numbers at level
3 as the cohort members age in France, in contrast to what we see in the UK.
The result is that as Cohort 2 members age between 19-21 and 31-33 there is
a 24 percentage point increase in the proportion at level 3 or above in France
(from 25% to 49%), compared to a 14 percentage point rise in the UK (from
30% to 44%). Therefore, there is more upgrading within Cohort 2 at all levels
(low and high qualifications) in France than in the UK (with most occurring in
France between the early and mid-twenties again).
Turning to the women in Cohort 2, Figure 14 below shows a similar situation
to their male counterparts. Thus we again observe greater improvement in
Cohort 2 relative to Cohort 1 for French women than for UK women. For
example, the proportion qualified at above level 3 by the age of 31-33 (29-31
in France) rises by 11 percentage points between Cohorts 1 and 2 in France
(from 26% to 37%) compared to 6 percentage points in the UK (from 24% to
30%).
29
0.5
0.4
0.3
0.2
0.1
0.0
1991 (age 19-21)
1997 (age 25-27)
2003 (age 31-33)
-o
<l
ev
el
le v 2
f w el
hi 2
ch
ac
vo
- o le v c
f w el
hi 3
ch
ac
v
>l oc
- o ev
f w el
hi 3
ch
ac
vo
c
proportion
Table 14: Cohort 2 - Females (France)
It is true that the progress within Cohort 2 is less for French women than for
French men. For example, the proportion failing to reach level 2 only falls by
7 percentage points in France (from 24% to 17%), compared to a fall of 8
percentage points in the UK (from 50% to 42%). The within cohort upgrading
advantage amongst women that France had in Cohort 1 therefore seems to
have disappeared by Cohort 2, at least at the lowest end of the distribution. It
remains true, however, that the proportion qualified at best at level 2 in France
falls significantly more as the cohort members age, and thus more women
reach higher qualification levels, than in the UK where this proportion barely
changes within Cohort 2.
The pooled results are in Figure 15.
30
0.5
0.4
0.3
0.2
0.1
0.0
1991 (age 19-21)
1997 (age 25-27)
2003 (age 31-33)
<l
ev
el
2
l
- o ev
e
fw l
hi 2
ch
ac
vo
c
l
- o ev
e
fw l
hi 3
ch
ac
v
> oc
- o lev
f w el
hi 3
ch
ac
vo
c
proportion
Figure 15: Cohort 2 - All (France)
Cohort 3
For the men in the French Cohort 33, Figure 16 below reveals a further fall in
the proportion failing to reach level 2 by the age of 19-21, down a further 9
percentage points, to 22%, compared to Cohort 2.
0.5
0.4
0.3
0.2
0.1
0.0
1997 (age 19-21)
2003 (age 25-27)
<l
ev
el
2
- o leve
fw l2
hi
ch
ac
vo
c
- o lev
f w el 3
hi
ch
ac
vo
>l c
- o ev
f w el 3
hi
ch
ac
vo
c
proportion
Figure 16: Cohort 3 - Males (France)
This figure is again less than the UK improvement discussed above and
shown in Figure 7, where there was a 17 percentage point fall in the
proportion failing to reach level 2 by the age of 19-21 between Cohorts 2 and
3, from 50% to 33%. The UK therefore continues to close the gap on France
in terms of the proportion of young people being qualified below level 2,
3
Again, the data end in 2001 in France, and so the second cohort are aged only 23-25 when
they are observed for the second time.
31
essentially through between cohort changes, and the introduction to the
workforce of more qualified workers.
However, further up the skills
distribution, there are again larger between cohort changes in France than in
the UK, as was observed between Cohorts 1 and 2. In particular, between
Cohorts 2 and 3 there is an 11 percentage point increase in the proportion of
19-21 year olds reaching at least level 3 in France (from 25% to 36%),
compared to a 8 percentage point increase in the UK (from 30% to 38%)
Looking within Cohort 3, even in this recent cohort, the UK is still not
upgrading qualifications as quickly as France. Thus, between the ages of 1921 and 25-27, the members of the French Cohort 3 still reduce the proportions
below level 2 (by 8 percentage points) and at level 2 (by 16 percentage
points), by more than the UK Cohort 3 (with reductions of 7 percentage points
and 10 percentage points respectively for below level 2 and exactly level 2).
In addition, at level 3 and above, Cohort 3 in France are upskilling very
rapidly, by 24 percentage points (from 36% to 60%), compared to 18
percentage points in the UK (from 38% to 56%).
The picture for the women in Cohort 3 is presented in Figure 17.
0.5
0.4
0.3
0.2
0.1
0.0
1997 (age 19-21)
2003 (age 25-27)
-o
<l
ev
el
lev 2
f w el
hi 2
ch
ac
vo
l
- o ev c
f w el
hi 3
ch
ac
vo
>
l
- o ev c
f w el
hi 3
ch
ac
vo
c
proportion
Figure 17: Cohort 3 - Females (France)
The first thing to note is that the proportion qualified below level 2 by the age
32
of 19-21 is again lower in Cohort 3 than in Cohort 2 (by 7 percentage points),
though like the men this is far smaller than the UK reduction of 20 percentage
points. However, as was the case for men, the proportion of French women
at level 3 or above by the age of 19-21 increases more between Cohorts 2
and 3 than in the UK, and then upgrades more quickly with age as well.
Finally for France, Figure 18 gives the pooled results.
0.5
0.4
0.3
0.2
0.1
0.0
1997 (age 19-21)
2003 (age 25-27)
<le
v
el
2
l
- o ev
e
fw l
h ic 2
ha
c
vo
- o lev c
f w el
hi 3
ch
ac
vo
c
>
l
- o ev
f w el
h ic 3
ha
c
vo
c
proportion
Figure 18: Cohort 3 - All (France)
In conclusion, then, for the UK-France comparison, the rate of upgrading from
below 2 within existing cohorts in the labour force seems to be faster in
France than in the UK between the ages of early and mid-twenties, although
this effect is then offset by greater upgrading from below level 2 between the
ages of mid-twenties and early-thirties in the UK, whereas little upgrading
takes place between these ages in France.
Higher up, however, even
allowing for the longer period of upgrading in the UK into the early thirties,
upgrading within cohorts in France to level 3 or beyond is substantially larger
than in the UK. This upgrading within cohorts in France is observed in all
three cohorts. Once a cohort of individuals are past the age of 18, therefore,
they are generally more likely to raise their qualifications in France than in the
UK. Where the UK is closing the qualifications gap on France is in terms of
33
upgrading between cohorts of the proportion reaching level 2, in the sense
that successive cohorts of individuals entering the labour market have larger
reductions in the numbers below level 2 in the UK than in France. However,
at level 3 and above, each successive cohort in France is reaching such a
level in increasing numbers, to a greater extent than the between cohort
changes in the UK.
Germany
Cohort 2
Unfortunately no data on a consistent basic were available for 1985 from
Germany, and so Cohort 1, who were aged 19-21 in 1985, could not be
studied.
We therefore start the analysis of qualifications upgrading in
Germany with an examination of the men in Cohort 2 (i.e. aged 19-21 in
1991). Their situation is depicted in Figure 19 below.
First, we note that the proportion of Cohort 2 members failing to reach level 2
by the age of 19-21 is at a similar level in Germany (33%) as in France (31%),
and so is again much lower than in the UK for Cohort 2 (50%). Despite this
smaller base to work with, however, Germany (as did France) achieves more
upgrading as this cohort age. Thus, between the ages of 19-21 and 31-33,
the proportion of this cohort failing to reach level 2 falls by 19 percentage
points (from 33% to 14%), compared to 12 percentage points in the UK (from
50% to 38%). As Figure 19 makes clear, all of this improvement in Germany
occurs by the time the initial low-achievers are aged in their mid-twenties.
34
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1991 (age 19-21)
1997 (age 25-27)
2003 (age 31-33)
<l
ev
el
- o lev 2
f w el
hi 2
ch
ac
vo
- o lev c
f w el
hi 3
ch
ac
vo
- o >lev c
f w el
hi 3
ch
ac
vo
c
proportion
Figure 19: Cohort 2 - Males (Germany)
Figure 19 also reveals a declining proportion at level 2 as the cohort members
age, and thus rising numbers at level 3 or even higher. As is well known for
Germany, the rising proportion at level 3 is caused mainly by people
completing apprenticeships, and this process mostly occurs by the midtwenties. The growth in level 4 or above qualifications is mostly, though not
exclusively, in terms of academic qualifications (degrees), and much of these
are acquired between the ages of 25-27 and 31-33. The total increase in the
size of the ‘level 3 or above’ group between the ages of 19-21 and 31-33 is 28
percentage points (from 51% of the cohort to 79% of the cohort). The size of
this increase is double the rate of increase of Cohort 2 men achieving at least
level 3 in the UK, which was 14 percentage points (from 30% to 44% at level
3 above as the cohort age). Thus, the extent of qualifications upgrading
within Cohort 2 is much greater in Germany than in the UK for men, at all
points of the qualifications distribution.
Similar findings are obtained for women, the position of whom in Cohort 2 in
Germany is shown in Figure 20 below. Thus the proportion of 19-21 year olds
failing to reach level 2 is significantly lower in Germany (32%) than in the UK
(50%), and then the proportion falls even further as the cohort members age
in Germany (by 15 percentage points to 17%) than in the UK (12 percentage
35
points to 38%). As Figure 20 makes clear, all of this reduction in the numbers
below level 2 occurs by the age of 25-27 in Germany, and therefore more
quickly than in the UK.
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1991 (age 19-21)
1997 (age 25-27)
voc
- of which ac
>level 3
voc
- of which ac
level 3
voc
- of which ac
level 2
2003 (age 31-33)
<level 2
Proportion
Figure 20: Cohort 2 - Females (Germany)
Figure 20 also makes plain the growth in qualifications at level 3 (before the
age of 25-27) and then above level 3 (by the age of 31-33) in Germany. This
growth in the proportion qualified at least to level 3 within Cohort 2 as the
cohort members age is again much larger for German women than for UK
women (a 25 percentage point rise from 50% to 75% at level 3+ between the
ages of 19-31 and 31-33 in Germany, compared to a 14 percentage point rise
from 28% to 42% in the UK). Thus, as for men, the extent of qualifications
upgrading within Cohort 2 as the women get older is significantly larger in
Germany than in the UK, at all points of the qualifications distribution.
Figure 21 shows the results for both genders together in the German Cohort
2.
36
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1991 (age 19-21)
1997 (age 25-27)
2003 (age 31-33)
<le
v
el
2
l
- o eve
fw l2
h ic
ha
c
vo
c
- o lev
f w el 3
h ic
ha
c
vo
c
- o >lev
f w el
h ic 3
ha
c
vo
c
proportion
Figure 21: Cohort 2 - All (Germany)
Cohort 3
The picture for German men in Cohort 3 is presented in Figure 22 below. The
first thing to note is that it is almost identical to the graph for German men in
Cohort 2 presented in Figure 19 above (except of course that Cohort 3 is only
observed twice). Thus the proportions at each qualification level at each point
in the cohort members’ lives are very similar in each cohort. Indeed, there is
a slightly higher proportion of Cohort 3 men failing to reach level 2 by the age
of 19-21 (35%) than was observed in Cohort 2 (33%). Similarly, a slightly
lower proportion of Cohort 3 reach at least level 3 by the age of 25-27 (78%)
than was observed in Cohort 2 at this age (79%). This contrasts with changes
described in the UK, above, where there was a 17 percentage point fall in the
proportion of 19-21 year olds failing to reach level 2 between Cohorts 2 and 3,
as well a 15 percentage point rise in the proportion of 25-27 year olds
reaching at least level 3. Therefore, across the two cohorts observed in both
the UK and Germany, there is no between cohort improvement in qualification
attainment in Germany and substantial improvement in the UK.
37
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1997 (age 19-21)
2003 (age 25-27)
<le
ve
l2
- o lev
f w el 2
h ic
ha
c
vo
c
- o leve
fw l3
h ic
ha
c
vo
- o >lev c
f w el 3
h ic
ha
c
vo
c
proportion
Figure 22: Cohort 3 - Males (Germany)
Within cohorts, in Cohort 3 there is still more upgrading in Germany than in
the UK, at all levels. Thus, as the members of Cohort 3 age from 19-21 to 2527, in Germany the proportion of men failing to reach level 2 falls by 20
percentage points and the proportion reaching at least level 3 rises by 30
percentage points. In the UK, the fall in the proportion failing to reach level 2
is just 7 percentage points, and the increase in the proportion reaching at
least level 3 is 18 percentage points.
A similar story can be told for the women of Cohort 3 in Germany. Thus there
is, very slightly, a higher proportion of Cohort 3 failing to reach level 2 by the
age of 19-21, compared to Cohort 2, and a slightly smaller proportion
reaching at least level 3 by the age of 25-27. This again contrasts with the
significant ‘between cohort’ progress observed in the UK between Cohorts 2
and 3. However, within Cohort 3, there remains more upgrading in Germany.
Thus, as Cohort 3 members ages from 19-21 to 25-27, the proportion failing
to reach level 2 falls by 16 percentage points in Germany and just 3
percentage points in the UK, while the proportion reaching at least level 3
rises by 26 percentage points in Germany and 14 percentage points in the
UK.
38
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1997 (age 19-21)
voc
- of which ac
>level 3
voc
- of which ac
level 3
voc
- of which ac
level 2
2003 (age 25-27)
<level 2
proportion
Figure 23: Cohort 3 - Females (Germany)
Finally for Germany, Figure 24 shows the situation for all members of Cohort
3.
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1997 (age 19-21)
2003 (age 25-27)
<l
ev
el
2
l
- o eve
fw l2
hi
ch
ac
vo
c
l
e
-o
v
e
fw l3
hi
ch
ac
v
> oc
- o le v
f w el 3
hi
ch
ac
vo
c
proportion
Figure 24: Cohort 3 - All (Germany)
In conclusion for the Anglo-German comparisons then, the UK lags Germany
in terms of the stock of skills by a considerable amount. In trying to close this
gap, the UK is not making any progress, and indeed the gap is widening,
when we look at upgrading the qualifications of particular cohorts of
39
individuals aged in their twenties. Thus, the proportion well-qualified simply
grows faster in Germany than in the UK within a given cohort, as the
members get older. The only way the UK is currently closing the gap on
Germany is via between cohort effects. Thus, in the UK, each cohort of
individuals who enter the labour market are progressively more well-qualified
than the previous one, especially amongst recent cohorts, and so they
increase the stock of qualifications in the labour force as they displace older,
less well-qualified individuals. In Germany, across the two cohorts studied,
there was no such effect, with both displaying very similar qualification levels.
USA
Unfortunately data on a consistent basis on qualifications in the US is only
available from 1994.
Therefore only Cohort 3 could be studied in the
American context, being first observed in 1997, at the age of 19-21.
In
addition, the American qualifications system does not allow an obvious split
into academic and vocational qualifications, and so such a distinction was
ignored. Figure 25 below shows the situation for US men in Cohort 3.
Figure 25: Cohort 3 - Males (US)
0.5
proportion
0.4
0.3
1997 (age 19-21)
0.2
2003 (age 25-27)
0.1
0.0
<level 2
level 2
level 3
>level 3
The proportion initially failing to reach level 2 at the age of 19-21 is 39%,
somewhat above the equivalent figure of 33% for the UK. However, there is
40
also more progress made in reducing this number in the US, since the
proportion falls by 9 percentage points to 30% by the time the cohort
members are aged 25-27. This compares to a 7 percentage point reduction in
the UK.
It should also be noted, however, in Figure 25, that the proportion at level 2
does not change too much as the cohort members age, whereas this
proportion fell in the UK amongst Cohort 3. This suggests more upgrading in
the UK than the US at higher levels, and the figures bear this out. As the
cohort members age from 19-21 to 25-27, the proportion reaching level 3 or
above rises by 10 percentage points in the US, and by 18 percentage points
in the UK.
A similar outcome is observed for women in the US, as depicted in Figure 26
below.
Figure 26: Cohort 3 - Females (US)
0.5
proportion
0.4
0.3
1997 (age 19-21)
0.2
2003 (age 25-27)
0.1
0.0
<level 2
level 2
level 3
>level 3
Thus, again there is a slightly higher proportion of Cohort 3 members initially
failing to reach level 2 at the age of 19-21 in the US than the UK (33% versus
30%). However, the US group make more progress as they move into their
mid-twenties, so that by the age of 25-27, the proportion of Cohort 3 failing to
reach level 2 is virtually the same in both countries.
Above this level,
41
however, there is more qualifications upgrading in the UK than the US as the
cohort members get older, with the proportion reaching at least level 3, for
example, rising by 14 percentage points in the UK for this cohort between the
ages of 19-21 and 25-27, and by 9 percentage points in the US.
Figure 27 shows the position for the pooled sample of men and women.
Figure 27: Cohort 3 - All (US)
0.5
proportion
0.4
0.3
1997 (age 19-21)
0.2
2003 (age 25-27)
0.1
0.0
<level 2
level 2
level 3
>level 3
In conclusion from the limited US evidence available, it seems that the skills
profile is quite similar in the US and the UK. Within cohorts there is slightly
more upgrading in the US at the lower end of the qualification distribution, but
significantly more upgrading in the UK at the higher end.
Conclusions
The main part of the report for the Skills Audit update considered the stocks of
skills in the five countries considered.
Comparisons with previous Skills
Audits then allowed a calculation of how these stocks are changing over time.
The general conclusion was that qualifications were being held by more
people at higher levels in each country. What a simple consideration of the
stocks could not tell us, however, is to what extent these improvements in
qualification levels were a result of upskilling amongst the existing workforce,
and to what extent they were a result of new, better-qualified cohorts entering
42
the workforce, at the expense of older less well-qualified cohorts. To give a
definitive answer to this question would require longitudinal data that allowed
the following of cohorts of individuals over time.
In the absence of such
suitable data in each country, this report therefore relied upon the construction
of pseudo cohorts from repeated cross-sectional data sources, to build up a
representative picture of the true cohorts of individuals.
The first part of this report revealed that there is significant qualifications
upgrading in the UK both within and between cohorts. In other words, each
group of individuals who join the labour force are better educated than the
previous group of joiners (between cohort changes), but also then acquire
more further qualifications as they age (within cohort changes, with the cut-off
point being age 31-33 in this report). Both aspects of upskilling, within and
between cohorts, help explain the rising skill profile of UK workers.
The main Skills Audit report revealed that the UK is lagging behind other
countries, particularly France and Germany, in terms of qualifications held by
adults. A comparison of the within and between cohort effects in the UK with
those in the other countries would reveal whether the UK is closing this skills
gap, and the principal route used in doing so. This was the focus of the
second half of this report. Figures 28 and 29 below summarise the findings.
They facilitate comparisons across countries, focussing on two proportions in
particular, respectively the proportion reaching level 2 or above, and the
proportion reaching the level 3 and above. The former proportion was chosen
for attention since it reveals the size of the movements out of low skills,
defined as below level 2. Similarly, level 3 and above is of interest, since it
often described by commentators as the level of skills now required of
individuals to perform successively in the modern world. All four countries are
shown on each diagram, with separate lines for each cohort. Changes within
cohorts are therefore represented by movements along one of these lines,
while changes between cohorts are represented by vertical distances
between cohort lines of the same country. For these diagrams, the data for
43
males and females are pooled4.
4
As usual, the data used to construct the figures can be found in Appendix B, in Tables A5
and A6 respectively.
44
Figure 28: Proportion at Level 2 or Above: Comparision Across Countries
1.0
0.9
France cohort 3
0.8
France cohort 2
p ro p o rtio n
0.7
UK cohort 3
0.6
Germany cohort 2
0.5
Germany cohort 3
0.4
US cohort 3
0.3
France cohort 1
UK cohort 2
0.2
UK cohort 1
0.1
0.0
19-21
25-27
31-33
age group
45
Looking firstly at the proportion reaching level 2 or above in Figure 28, we can
see that the slope of the lines, representing the size of within cohorts
changes, is certainly steepest in Germany, revealing the largest amount of
movement out of the below level 2 group in that country. The slope of the
lines in the other three countries are not too different, however, suggesting
upgrading up to at least level 2 is similar in France and the US to the situation
in the UK. Indeed, between the ages of 25-27 and 31-33, where data is
available it suggests no upgrading from below level 2 in any country but the
UK (although of course this is largely a consequence of more individuals
requiring upgrading in the UK).
Comparing the vertical shifts of cohort lines across countries, it is clearly the
case that the largest movement between cohorts in the proportion qualified to
level 2 or above has been in the UK.
The lines for Cohorts 2 and 3 in
Germany are virtually on top of each other, suggesting no change across
cohorts, while the shift up of cohort lines is less in France than in the UK.
Thus, for Cohorts 1 and 2 in the UK, the proportion qualified to level 2 or
above is clearly below the lines for all other cohorts in all other countries,
while for Cohort 3, the UK line has shifted up to such an extent that it is very
much within the cluster of lines for the other countries. Figure 28 therefore
makes clear how between cohort changes in particular have helped the UK
close the gap on the other countries with respect to the existence of low skills
(individuals below level 2).
Figure 29 performs a similar analysis for the proportions at level 3 and above.
Comparing the slopes of lines shows that at this higher level, the extent of
upgrading within cohorts is certainly less in the UK (shallower lines) than in
both France and Germany. The detailed results presented above showed
that the greater upgrading in the continental European countries compared to
the UK was due largely to the greater use of the vocational route to level 3
and above in the former countries, particularly in Germany (due, of course, to
the apprenticeship system). It should also be pointed out, though, that as at
level 2 and above, the only evidence of upgrading to level 3 or above between
the ages of 25-27 and 31-33 in any country is in the UK.
46
Figure 29: Proportion at Level 3 or Above: Comparison Across Countries
0.9
p ro p o rtio n
0.8
Germany cohort 2
0.7
Germany cohort 3
0.6
US cohort 3
France cohort 3
0.5
UK cohort 3
0.4
France cohort 2
0.3
UK cohort 2
0.2
UK cohort 1
France cohort 1
0.1
0.0
19-21
25-27
31-33
age group
47
Comparing between cohort shifts across countries, again for Germany we see
a zero change. Looking just at France and the UK, it is clear that although
there has been between cohort improvement in the UK, this is less than such
improvement in France. These lower relative between cohort shifts in the UK
at level 3 or above, coupled with the much lower within cohorts shifts
compared to other countries, means that the UK line even for Cohort 3 is still
noticeably below the Cohort 3 lines for the other countries, particularly
Germany, and so the UK has more to do to close the skills gap on other
countries at level 3 and above.
Figures 30 through to 33 present the same results in a different way, and
directly display the size of the within and between cohort changes as bars in a
bar chart, to make it easier to compares size of effects. Figure 30 shows the
size of within cohort changes at level 2 or above. Looking at changes just
between the ages of 19-21 and 25-27, these are smaller in the UK for every
cohort than in all other countries, although it is true that the greater upgrading
between the ages of 25-27 and 31-33 in the UK means the total upgrading
across the two time periods is actually the largest in the UK.
Figure 30: Within Cohort Changes, Level 2 and Above
0.25
0.2
(31-33) - (25-27)
0.1
(25-27) - (19-21)
0.05
Fr
an
3
ce
co
ho
Fr
rt
an
1
ce
co
ho
Fr
rt
an
2
ce
co
Ge
ho
rm
rt
an
3
yc
oh
Ge
or
rm
t2
an
yc
oh
or
t3
US
co
ho
rt
3
2
co
ho
rt
UK
UK
co
ho
rt
-0.05
co
ho
rt
1
0
UK
Proportion
0.15
48
Figure 31 shows the between cohort changes at level 2 and above, and make
clear the improvement the UK has made in this regard, having larger between
cohort changes than France, particularly between Cohorts 2 and 3, and of
course also larger changes than Germany, where the between cohort change
is tiny and negative.
Figure 31: Between Cohort Changes, Level 2 and Above
0.2
proportion
0.15
0.1
0.05
0
UK cohort 2 cohort1
-0.05
UK cohort 3cohort 2
France cohort 2 - France cohort 3 - Germany cohort
cohort 1
cohort 2
3-cohort 2
At level 3 or above, the within cohort gains are much less in the UK than in
Germany or France, even when the greater upgrading between the ages of
25-27 and 31-33 is added in, as shown in Figure 32.
Figure 32: Within Cohort Changes, Level 3 and Above
0.25
0.2
0.15
(31-33) - (25-27)
0.1
(25-27) - (19-21)
Fr
an
2
ce
c
oh
Ge
ort
rm
3
an
yc
oh
Ge
ort
rm
2
an
yc
oh
ort
3
US
co
ho
rt
3
1
Fr
an
ce
co
ho
rt
3
Fr
an
ce
co
ho
rt
2
co
ho
rt
UK
co
ho
rt
UK
co
ho
rt
-0.05
1
0.05
0
UK
proportion
0.3
49
Finally, Figure 33 shows the size of the between cohort shifts at level 3 and
above, which are clearly larger in France than in the UK, suggesting that
France is closing the level 3 gap on Germany (where the between cohort level
3+ shifts are again tiny and negative) more quickly than the UK.
Figure 33: Between Cohort Changes, Level 3 and Above
0.25
0.2
proportion
0.15
0.1
0.05
0
-0.05
UK cohort 2 cohort1
UK cohort 3cohort 2
France cohort 2 - France cohort 3 - Germany cohort
cohort 1
cohort 2
3-cohort 2
In summary, then, although the education level of young people in the UK is
rising (both within and between cohorts), in only some respects are gains
being made on other countries, in terms of closing the skills gap identified in
the main Skills Audit report. Most of the relative improvement in the UK’s
position is occurring at the lower end of the qualifications distribution, in terms
of reducing the numbers failing to reach level 2. In particular, the reduction in
the numbers with such low skills in successive cohorts is greater in the UK
than in either France or Germany (the between cohort effect in Germany
being essentially zero). As each cohort ages, further numbers also move out
of the below level 2 group. Such within cohort upgrading occurs to a similar
extent in France and the UK, but both trail Germany. Higher up the skills
distribution, the UK is making even fewer gains. Both within and between
cohort changes in the proportion qualified to level 3 and above are higher in
France than in the UK. Within cohort changes at this level in Germany are
50
higher than in both countries, due in large part to level 3 vocational
qualifications obtained through apprenticeship, although the between cohort
changes are again zero at this level in Germany, essentially because of the
already high numbers reaching level 3 in earlier cohorts. In terms of level 3+
skills, therefore, French young people are pulling away from those in the UK,
and closing the gap on Germany more quickly.
Thus, almost all of the relative gains that the UK is making is in terms of
between cohorts changes rather than within cohort changes. Of course, given
the relative starting points of the countries, it would have been disappointing if
the UK had not made larger gains between cohorts than France and
Germany, since there was more room for improvement in the UK. Indeed, the
German results show no improvement in attainment between Cohorts 2 and
3. Of course, if a country is at or close to achieving its maximum potential
from its human resources, then such an outcome would be expected, with
each cohort at this maximum. This is what the UK must aspire to.
The results also suggest that even if the UK continued to make improvements
between cohorts, with each new cohort entering the workforce becoming
successively better qualified, and this continued until the UK was matching
the starting qualification levels of new cohorts in France and Germany, if the
existing workforce are not being upgraded, it will take many years before the
new cohorts work through the system and fully replace the existing lower
qualified cohorts in the workforce.
Even once this is achieved, without
improvements within cohorts as cohort members age, then even if the UK can
match the starting qualifications of France and Germany, a gap would simply
open up again as the latter two countries upgraded their workers further as
they grow older. Of course, it might be the case that within cohort effects will
follow from between cohort effects, and that once new entrants to the UK
workforce are more highly qualified on average, then they will be in a position
to go on acquiring more qualifications and so upgrade qualifications within
cohorts like the other two countries.
Without such improvements in
continuous upgrading within cohorts, the UK is never going to catch France
and Germany, unless the UK’s between cohort effects become implausibly
51
large.
The improvements in successive cohorts in the UK are therefore
promising and welcome, and will help close the skills gap on the other
countries, but more needs to be done to continuously upgrade these cohorts
as they grow older, if the gap is to be closed further.
52
Appendix A: Allocation of Qualifications to UK National Qualification
Framework (NQF) Levels
The UK National Qualifications Framework
For information on the UK National Qualifications Framework and a description of
the Framework levels go to
http://www.qca.org.uk/qualifications/types/493.html
Allocation of UK qualifications to UK NQF levels
For the purposes of international comparisons, UK qualifications have not been
allocated in the same proportions as in the NQF framework . Instead, allocations
reflect the need to achieve equivalence with qualifications in the other countries in the
audit.
The qualifications assigned to the four Levels using the 1985 Labour Force
Survey ordering are shown below. Qualifications in italics classified as
vocational.
> Level 3
All allocations 100% unless otherwise stated.
Higher degree
First degree
Other degree
BTEC HNC/HND
Secondary Teaching
Primary Teaching
Nursing
Level 3
All allocations 100% unless otherwise stated.
BTEC ONC/OND
City & Guilds 20%
One or more A-level 80%
(G) NVQ 3 (from 1996)
Other qualifications 10%
Level 2
All allocations 100% unless otherwise stated.
City & Guilds qualification 40%
One or more A-level 20 %
(G) NVQ 2 (from 1996 onwards)
One or more O level, GCSE grade A-C or equivalent 40%5
5
40% of O level passes / GCSE grade A-C qualifications are allocated to level 2, since this is
an estimate of the proportion of those with at least one O level / GCSE as their highest
qualification who obtain an O level/GCSE pass in English, Maths and one other subject. This
makes the UK level 2 compatible with that in other countries, where such passes in maths,
national language and one other subject are required to reach level 2.
53
Other qualifications 35%
< Level 2
All allocations 100% unless otherwise stated.
City & Guilds qualification 40%
Apprenticeship without recognised vocational qualification
One or more O level, GCSE grade A-C or equivalent 60%
CSE < Grade 1
Other qualifications 55%
No qualifications
Don’t Know
Allocation of UK qualifications using the 1996 Labour Force Survey
Ordering
This ordering used for Cohort 3, observed only after the change in the
classification in 1996. Qualifications in italics classified as vocational.
> Level 3
All allocations 100% unless otherwise stated.
Higher degree
NVQ level 5
First degree
Other degree
NVQ Level 4
Diploma in higher education
HNC,HND,BTEC etc higher
Teaching, further education
Teaching, secondary education
Teaching, primary education
Teaching, level not stated
Nursing etc
RSA higher diploma
Other HE below degree
Level 3
All allocations 100% unless otherwise stated.
NVQ level 3
GNVQ advanced
One or more A level or equivalent 70%
RSA advanced diploma
OND,ONC,BTEC etc, national
City & Guilds advanced craft
Scottish CSYS 67%
SCE higher or equivalent 50%
One or more AS level or equivalent 17%
Other Qualifications 10%
54
Level 2
All allocations 100% unless otherwise stated.
One or more A level or equivalent 30%
Scottish CSYS 33%
SCE higher or equivalent 50%
One or more AS level or equivalent 26%
Trade apprenticeship 51%
NVQ level 2
GNVQ intermediate
RSA diploma
City & Guilds craft
BTEC,SCOTVEC first or general diploma
One or more O level, GCSE grade A-C or equivalent 40%
Other Qualifications 35%
< Level 2
All allocations 100% unless otherwise stated.
One or more AS level or equivalent 57%
Trade apprenticeship 49%
One or more O level, GCSE grade A-C or equivalent 60%
NVQ level 1
GNVQ,GSVQ foundation level
One or more CSE below grade1,GCSE below grade C
BTEC,SCOTVEC first or general certificate
SCOTVEC modules
RSA other
City & Guilds other
YT,YTP certificate
Other Qualifications 55%
No qualifications
Don't know
Allocation of French qualifications to NQF qualification levels
Qualifications in italics classified as vocational
> Level 3
All allocations 100% unless otherwise stated.
All degrees and higher degrees
BTS, DUT etc. Qualifications requiring two years of study after the
Baccalaureat
Level 3
All allocations 100% unless otherwise stated.
General Baccalaureat
Technical Baccalaureat
Vocational Baccalaureat
55
Vocational Brevet
Level 2
All allocations 100% unless otherwise stated.
Brevet 89%
CAP/BEP
< Level 2
All allocations 100% unless otherwise stated.
Don't Know
No qualification or only Certificate of Primary Education (CEP)
Brevet 11%
Allocation of German qualifications to NQF qualification levels
Qualifications in italics classified as vocational
>Level 3
All allocations 100% unless otherwise stated.
Degree or higher degree from University
Degree qualification from Fachhochschule
Meister or Techniker certificate or Fachschulabschluss
Level 3
All allocations 100% unless otherwise stated.
Hochschulreife (Qualification giving right to enter university or
Fachhochschule )
Fachhochschulreife (Qualification giving right to enter Fachhochschule )
Apprenticeship with prior school leaving qualification 94%
Level 2
All allocations 100% unless otherwise stated.
Leaving School Certificate of the Realschule
Apprenticeship with leaving certificate from the Polytechnische Oberschule of
the DDR (Germany 2002 only)
Apprenticeship with prior school leaving qualification 6%
Apprenticeship without prior school leaving qualification
Level 1
All allocations 100% unless otherwise stated.
Leaving Certificate of the Polytechnische Oberschule of the DDR ( Germany
2002 only)
Leaving Certificate of the Secondary Modern School (Hauptschulabschluss)
Others (includes Don’t Know, No Response and No Qualifications)
Allocation of US qualifications to NQF qualification levels
All allocations 100% unless otherwise stated.
56
>Level 3
Bachelor or higher degree including professional
Associate degree
Level 3
Some college
Level 2
High School graduation only 50%
<Level 2
High School graduation only 50%
Less than High School graduation
57
Appendix B: Figures Used for All Graphs in Report
Table A1: Proportion of Cohort Members at Each Qualification Level (UK
– Figures 1-9)
Cohort 1
<level 2
level 2
- of which
ac
- voc
level 3
- of which
ac
- voc
>level 3
- of which
ac
- voc
Cohort 2
<level 2
level 2
- of which
ac
- voc
level 3
- of which
ac
- voc
>level 3
- of which
ac
- voc
Cohort 3
<level 2
level 2
- of which
ac
- voc
level 3
- of which
ac
- voc
>level 3
- of which
ac
- voc
1985
0.64
0.17
0.12
0.05
0.17
0.10
Males
1991
0.51
0.15
0.09
0.06
0.15
0.07
1997
0.41
0.19
0.11
0.08
0.14
0.06
1985
0.64
0.19
0.17
0.02
0.14
0.10
Females
1991 1997
0.53
0.46
0.17
0.19
0.15
0.15
0.02
0.13
0.08
0.04
0.11
0.07
1985
0.64
0.18
All
1991
0.52
0.16
1997
0.44
0.19
0.15
0.03
0.15
0.12
0.04
0.14
0.13
0.06
0.12
0.10
0.05
0.03
0.08
0.07
0.17
0.07
0.06
0.25
0.06
0.03
0.01
0.08
0.18
0.13
0.07
0.26
0.17
0.04
0.03
0.01
0.05
0.17
0.10
0.04
0.24
0.14
0.02
1991
0.50
0.20
0.06
1997
0.40
0.19
0.09
2003
0.38
0.18
0.02
1991
0.50
0.22
0.07
1997
0.42
0.21
0.10
2003
0.38
0.20
0.01
0.02
1991
0.50
0.21
0.11
0.06
1997
0.41
0.20
0.15
0.10
2003
0.38
0.19
0.14
0.06
0.27
0.10
0.09
0.14
0.10
0.08
0.13
0.18
0.03
0.25
0.15
0.06
0.13
0.14
0.06
0.12
0.16
0.05
0.26
0.12
0.07
0.14
0.12
0.07
0.13
0.16
0.12
0.03
0.06
0.08
0.26
0.05
0.08
0.31
0.16
0.09
0.03
0.07
0.06
0.25
0.05
0.07
0.30
0.16
0.10
0.03
0.07
0.07
0.25
0.05
0.07
0.30
0.01
0.02
1997
0.33
0.29
0.19
0.07
2003
0.26
0.19
0.23
0.09
0.01
0.02
1997
0.30
0.32
0.18
0.07
2003
0.26
0.21
0.20
0.10
0.01
0.02
1997
0.31
0.30
0.18
0.07
2003
0.26
0.20
0.21
0.09
0.18
0.11
0.33
0.10
0.08
0.19
0.20
0.12
0.33
0.11
0.10
0.17
0.19
0.11
0.33
0.11
0.09
0.18
0.22
0.10
0.05
0.07
0.12
0.37
0.23
0.10
0.06
0.07
0.10
0.36
0.23
0.10
0.05
0.07
0.11
0.36
0.02
0.03
0.30
0.07
0.03
0.03
0.30
0.06
0.02
0.03
0.30
0.07
58
Table A2: Proportion of Cohort Members at Each Qualification Level
(France – Figures 10-18)
Cohort 1
<level 2
Level 2
- of which
ac
- voc
level 3
- of which
ac
- voc
>level 3
- of which
ac
- voc
Cohort 2
<level 2
level 2
- of which
ac
- voc
level 3
- of which
ac
- voc
>level 3
- of which
ac
- voc
Cohort 3
<level 2
level 2
- of which
ac
- voc
level 3
- of which
ac
- voc
>level 3
- of which
ac
- voc
Males
1985 1991
0.40
0.25
0.54
0.43
1997
0.24
0.43
Females
1985 1991 1997
0.37
0.24
0.23
0.50
0.37
0.36
1985
0.39
0.52
All
1991
0.25
0.40
1997
0.23
0.40
0.07
0.47
0.05
0.05
0.38
0.12
0.05
0.38
0.11
0.11
0.39
0.12
0.07
0.30
0.16
0.07
0.29
0.15
0.09
0.43
0.08
0.06
0.34
0.14
0.06
0.34
0.13
0.02
0.04
0.00
0.05
0.07
0.20
0.04
0.06
0.22
0.05
0.06
0.01
0.07
0.08
0.23
0.07
0.08
0.26
0.03
0.05
0.01
0.06
0.08
0.21
0.06
0.07
0.24
0.00
0.00
1991
0.31
0.44
0.12
0.08
1997
0.18
0.33
0.14
0.08
2003
0.18
0.33
0.00
0.01
1991
0.24
0.41
0.16
0.07
1997
0.17
0.27
0.19
0.07
2003
0.18
0.28
0.00
0.01
1991
0.27
0.43
0.14
0.07
1997
0.17
0.30
0.17
0.08
2003
0.18
0.30
0.19
0.25
0.22
0.04
0.28
0.19
0.05
0.28
0.16
0.21
0.20
0.31
0.04
0.23
0.20
0.04
0.24
0.17
0.20
0.22
0.27
0.04
0.26
0.19
0.04
0.26
0.17
0.16
0.06
0.03
0.07
0.12
0.31
0.06
0.10
0.33
0.24
0.07
0.04
0.09
0.11
0.36
0.07
0.10
0.37
0.20
0.07
0.03
0.08
0.11
0.34
0.07
0.10
0.35
0.02
0.01
1997
0.22
0.42
0.19
0.12
2003
0.14
0.26
0.20
0.13
0.03
0.01
1997
0.17
0.36
0.26
0.11
2003
0.11
0.20
0.25
0.11
0.02
0.01
1997
0.19
0.39
0.22
0.11
2003
0.13
0.23
0.23
0.12
0.19
0.23
0.33
0.04
0.22
0.29
0.19
0.17
0.43
0.05
0.16
0.27
0.19
0.20
0.38
0.04
0.19
0.28
0.26
0.07
0.03
0.13
0.16
0.31
0.36
0.07
0.05
0.14
0.12
0.42
0.31
0.07
0.04
0.14
0.14
0.36
0.02
0.01
0.18
0.13
0.04
0.01
0.29
0.13
0.03
0.01
0.24
0.13
59
Table A3: Proportion of Cohort Members at Each Qualification Level (Germany
– Figures 19-24)
Cohort 2
<level 2
level 2
- of which
ac
- voc
level 3
- of which
ac
- voc
>level 3
- of which
ac
- voc
Cohort 3
<level 2
level 2
- of which
ac
- voc
level 3
- of which
ac
- voc
>level 3
- of which
ac
- voc
Males
1991 1997
0.33
0.14
0.17
0.06
2003
0.14
0.06
Females
1991 1997 2003
0.32
0.17
0.18
0.18
0.07
0.07
1991
0.33
0.18
All
1997
0.16
0.07
2003
0.16
0.06
0.14
0.02
0.50
0.03
0.03
0.66
0.03
0.03
0.51
0.17
0.02
0.49
0.04
0.03
0.63
0.04
0.03
0.55
0.15
0.02
0.49
0.03
0.03
0.65
0.04
0.03
0.53
0.18
0.32
0.01
0.23
0.44
0.13
0.12
0.40
0.28
0.19
0.29
0.01
0.21
0.43
0.12
0.14
0.41
0.20
0.19
0.30
0.01
0.22
0.43
0.13
0.13
0.41
0.24
0.00
0.01
1997
0.35
0.16
0.08
0.05
2003
0.15
0.07
0.19
0.09
0.00
0.01
1997
0.33
0.18
0.09
0.04
2003
0.17
0.08
0.15
0.05
0.00
0.01
1997
0.34
0.17
0.09
0.04
2003
0.16
0.07
0.17
0.07
0.14
0.02
0.47
0.04
0.03
0.65
0.16
0.02
0.47
0.05
0.03
0.60
0.15
0.02
0.47
0.04
0.03
0.62
0.21
0.27
0.01
0.27
0.38
0.13
0.24
0.23
0.02
0.25
0.35
0.15
0.22
0.25
0.01
0.26
0.36
0.14
0.01
0.01
0.08
0.05
0.01
0.01
0.10
0.04
0.01
0.01
0.09
0.04
Table A4: Proportion of Cohort Members at Each Qualification Level (US
– Figures 25-27)
Cohort 3
<level 2
level 2
level 3
>level 3
1997
0.39
0.18
0.41
0.03
Males
2003
0.30
0.16
0.21
0.33
1997
0.33
0.17
0.47
0.04
Females
2003
0.26
0.14
0.21
0.39
1997
0.36
0.17
0.44
0.03
All
2003
0.28
0.15
0.21
0.36
60
Table A5: Proportion of All Cohort Members at Level 2 or Above, by
Cohort and Country (Figure 28)
Age
19-21
25-27
31-33
UK cohort 1
0.36
0.48
0.57
UK cohort 2
0.50
0.59
0.62
UK cohort 3
0.69
0.74
France cohort 1
0.61
0.75
0.77
France cohort 2
0.73
0.83
0.82
France cohort 3
0.81
0.87
Germany cohort
0.67
0.84
0.66
0.84
0.65
0.72
0.84
2
Germany cohort
3
US cohort 3
Table A6: Proportion of All Cohort Members at Level 3 or Above, by
Cohort and Country (Figure 29)
Age
19-21
25-27
31-33
UK cohort 1
0.19
0.32
0.38
UK cohort 2
0.29
0.39
0.43
UK cohort 3
0.38
0.54
France cohort 1
0.09
0.35
0.37
France cohort 2
0.30
0.53
0.52
France cohort 3
0.42
0.64
0.50
0.78
3
0.49
0.76
US cohort 3
0.47
0.57
Germany cohort
2
0.78
Germany cohort
61
Table A7: Within and Between Cohort Changes in the Proportion
Qualified at Level 2 or Above, All Cohort Members (Figures 30 and 31)
Age
(25-27)(19-21)
Cohort 1
France
Germany
0.116
0.141
0.085
0.011
0.090
0.099
0.169
0.031
-0.007
-0.004
0.054
0.065
0.178
0.136
0.114
0.186
0.081
US
(31-33)
–
(25-
27)
Within
(25-27)-
cohort
changes
UK
(19-21)
Cohort 2
(31-33)
–
(25-
27)
Cohort 3
(25-27)(19-21)
0.078
Cohort 2
Between
cohort
changes
– Cohort 19-21
1
Cohort 3
– Cohort 19-21
-0.015
2
62
Table A8: Within and Between Cohort Changes in the Proportion
Qualified at Level 3 or Above, All Cohort Members (Figures 32 and 33)
Age
(25-27)(19-21)
Cohort 1
France
Germany
0.132
0.260
0.059
0.018
0.099
0.228
0.278
0.040
-0.011
-0.001
0.161
0.222
0.276
0.107
0.208
0.089
0.121
US
(31-33)
–
(25-
27)
Within
(25-27)-
cohort
changes
UK
(19-21)
Cohort 2
(31-33)
–
(25-
27)
Cohort 3
(25-27)(19-21)
0.102
Cohort 2
Between
cohort
changes
– Cohort 19-21
1
Cohort 3
– Cohort 19-21
-0.011
2
63
Appendix C: Sample Sizes
Approximate size of an observed pseudo cohort in each country:
France: 8,000
Germany: 2,000
UK: 7,500
US: 5,000
The UK LFS Users Guide reports that for the analysis of small age groups of
3 years width, such as those used in this report, the standard error on
estimated proportions is about 1%, and thus the true proportion in the
population of this age may be +/- 2 percentage points around the estimated
figures presented here. With a similar sample size, the standard errors on
French estimates will be similar to this figure, while the US and the German
standard errors will be larger, given the proportionally smaller data sets
available for those countries.
64
Copies of this publication can be obtained from:
DfES Publications
P.O. Box 5050
Sherwood Park
Annesley
Nottingham
NG15 0DJ
Tel: 0845 60 222 60
Fax: 0845 60 333 60
Minicom: 0845 60 555 60
Oneline: www.dfespublications.gov.uk
© Centre for Economic Performance 2005
Produced by the Department for Education and Skills
ISBN 1 84478 419 3
Ref No: RR621
www.dfes.go.uk/research