2000, the German economy posted its most rapid growth rate since unification.
In the same year, sales of cars fell by some 11%. Both the onset of the decline,
and its duration – the weak market has persisted until at least the second quarter
of 2001 – took the industry by surprise. Here we set out some features of this
decline, and look at some hypotheses about its causes.
Who stopped buying?
The first issue to be looked at is: who has stopped buying new cars? The Kraftfahrt
Bundesamt (KBA) maintains a comprehensive database of information on vehicle
registrations, and this provides help in identifying the type of buyer that
went missing from the market when the decline set in. Two separate tabulations
by the KBA split the buyers into the types of enterprises buying the vehicles,
and also split the private buyers by age and sex. In round numbers, about 55%
of German purchases are by individuals whose age and sex are identified in the
KBA data. A further 20% are registered by the motor trade itself, and the remaining
25% are sold to a range of other enterprises and organisations.
It was the private buyers who were responsible for the decline in sales. Total
car sales fell by some 423,000 units in 2000 compared with the previous year,
and 96% of this decline (406,000 units) corresponded to the reduction in sales
to individuals whose age and sex are identified by the KBA. Sales to the motor
trade, and to other enterprises and organisations, continued at approximately
the same level.
Among the private buyers, the decline was apparent in every age group, but
it was the younger age groups who contributed disproportionately to the decline
in sales. In round numbers, about a quarter of new car purchases were by individuals
below the age of 45, but nearly half of the decline in purchases between 1999
and 2000 was due to reduced purchases by these younger buyers. If we look at
the distribution of purchases by sex, we find that the share of purchases by
female buyers was about the same in both years. The proportion of buyers that
are female tends to be higher in the younger age groups, and it seems that any
slight change in the distribution by sex was merely a reflection of the change
in the distribution by age group.
In terms of the geographical distribution of the decline, it was slightly steeper
in Eastern Germany than in Western Germany, with sales in the new Länder down
by about 17%, compared with a 10% decline in the Western Länder. However, the
relatively small importance of the East is apparent from the fact that the overall
decline in sales was so close to the decline in the West.
“It was the private buyers who
were responsible for the decline in sales.“
Used car purchases and vehicles in use
Why did these private buyers absent themselves from the market? One hypothesis
that has been heard in the German trade and industry is that the weakness of
new car sales has been due to the greater relative attraction of used vehicles.
Is there, then, evidence that these predominantly younger buyers were compensating
for buying fewer new vehicles by buying more used ones?
No, there is no such evidence. The KBA also tracks used car sales in the same
way. Younger buyers, not surprisingly, are much more important in the used car
market than in the market for new cars, but the pattern of increased reluctance
to buy on the part of younger buyers is even more apparent in the used car market
than in the new car market. Overall, used car sales were down by 4%, or 320,000
units, in 2000. Essentially all (97%) of this decline could be attributed to
buyers under the age of 35.
With fewer vehicles, whether new or used, being purchased, particularly by
younger buyers, was there some noticeable change in the extent of vehicle ownership
among these age groups? The KBA publishes data on vehicle ownership split by
the age of the owner, so this can also be tested.
Although they bought fewer vehicles, the extent of vehicle ownership continued
to rise, relative to the population, in each age group up to the age of 60.
There was a slight decline in the ratio of vehicles owned to population in the
age groups above 60, but this has nothing to do with the phenomenon of falling
car sales with which we are concerned. The absolute number of vehicles owned
by the 25-35 age group fell slightly, but this is more than offset by the decline
in the population in those age groups.
We have illustrated these features of the decline by a comparison between 1999
and 2000. The weakness of demand has persisted into 2001, and examination of
the same data for 2001 produces essentially the same conclusions, pinpointing
the reluctance of younger buyers to purchase either new or used vehicles as
the main feature of the decline.
The age distribution of the vehicle parc
The data is therefore telling us that, while private owners continue to possess
vehicles in the same or slightly growing numbers, they failed to replace those
vehicles, either with new or used ones, at the same rate in 2000 as they had
done in the previous year. Were there features of the age distribution of the
vehicles in use immediately prior to the sales decline, which could have led
to this? Again, the KBA’s tabulations of vehicles in use by age of vehicle can
be consulted in order to throw light on this question.
This, too, turns out to be a blind alley. The age distribution of the vehicle
parc is a complex thing, which is only roughly summarised by looking at the
average age of vehicles in use. On that measure, the average age was 6.76 years
in July 1999, just before the decline set in. This is very slightly below the
average for the recent past, (6.78 for the five years from 1995 to 2000), but
the difference is too slight to have any explanatory value as a factor contributing
to the sales decline. More significant may have been the fact that the proportion
of vehicles less than three years old, at just under 21.9% of all vehicles in
use, was above the five-year average of 21.2% – such vehicles can more readily
serve as substitutes for new cars.
However, if these small changes in the age distribution of the parc had any
influence on the demand for new cars, that influence would have faded by now.
By mid-2000, the average age had risen to 6.82, and the share of vehicles less
than three years old had fallen to 21.6%. It will be some time before the mid-2001
data are available, but since there is a high degree of predictability about
the scrapping rates by age, they can be forecast with some confidence. The average
age of the parc will have risen to about 6.9 years, and the proportion of the
parc under 3 years old will have fallen to 20.7%.
Do newer models mean more (total) sales?
It must be remembered that the high sales in 1999 were partly the result of
an abnormally large bank of unfilled orders that were carried through from 1998,
following launches of significant new versions of such models as the Golf and
Astra. Another hypothesis that is sometimes advanced is that it is the lack
of such new models that has caused the sales dearth.
A quick way of refuting this hypothesis (as well as other ones blaming the
market decline on the fuel price, or the relative price of new and used cars)
is to compare the evolution of the German market with other neighbouring markets
which underwent similar changes in fuel prices, and also in the nature of the
models on offer. These neighbouring markets did not suffer the same symptoms
that have been seen in Germany. A more formal examination of the “model-age”
hypothesis can be undertaken by measuring the age since launch of each model
on the market, and constructing a weighted average (which will be different
in each national market, depending on the market share of particular models).
This average will go down when important new models are launched, and, if the
hypothesis is correct, a reduction in the average age of the models on the market
should lead, other things being equal, to an increase in sales after a certain
time-lag (necessary for the models to move from launch to full commercial availability).
Testing this hypothesis for each European market produces disappointing results.
Although the market share of a particular vehicle depends strongly on its age,
the size of the total market appears relatively impervious to changes in the
average age since launch of all the models available.
“buyers of smaller vehicles revert
to buying used cars when times are less good.“
What vehicles did they stop buying?
The decline in German car sales between 1999 and 2000 hit all vehicle segments
with approximately equal force. There is just one exception to this statement:
sales in the executive segment inched ahead. Now this statement must be immediately
qualified by saying that the dividing line between the “upper medium” and “executive”
segments is even more arbitrary than most of the dividing lines between segments.
At the lower end of the executive segment, the (still relatively new) BMW 3
Series and the Mercedes C Class were still on the upward-sloping parts of their
product cycles, and stealing market share from vehicles in the upper medium
segment. If one combines the executive and the upper medium segment, total sales
in the two segments combined fell at exactly the same rate as the market as
a whole. The other group of vehicles whose sales surged were the mini-MPVs (such
as the Renault Scenic). However, we consider these as a sub-segment within the
lower medium segment, and the lower medium segment as a whole declined at a
rate slightly higher than the market average of 11%.
However, the strength of executive segment sales should not be dismissed purely
as a matter of definition. It makes sense in the light of the evidence we have
reviewed about the changes in the buyer profile: vehicles whose sales depend
more on corporate purchases, or sales to older individuals, have been less affected
than the vehicles which appeal to younger buyers. This theme has become much
more apparent in the segmentation trends during 2001, which have revealed a
stronger performance in both the executive and the upper medium segment (the
latter having been boosted by the replacement of the Mondeo and Laguna, and
the rejuvenation of the Passat); while the lower medium and small segments have
This tendency towards upsizing – or at least the absence of downsizing – in
the German market contrasts with what has been seen in the European markets
that have remained relatively strong. In most of these markets, the share of
smaller vehicles has tended to increase. This contrast supports an idea which
is often, though not universally true: namely that the buyers of larger vehicles
are more likely to remain in the market in good times or bad, while the buyers
of smaller vehicles come into the market for new cars when they are feeling
confident and wealthy, but revert to buying used cars when times are less good.
In the market for new cars (to misquote St Mark) “the rich you have always with
you, and whensoever you will, you may sell them cars, but the poor you have
not always with you”.
So far we have established facts about the particular buyers whose absence
from the market caused the downturn, but we have not explained that downturn
– we have only looked at, and in part rejected, certain hypotheses about the
reasons for the downturn. What those hypotheses had in common was that they
all sought to find explanations in factors that were specific to the automotive
sector. We say that we rejected them in part, because they fail to explain why
the same alleged cause did not produce a similar effect in different countries
at the same time, or in Germany at other times. However, these could still have
been contributory factors.
Economic factors must be the other part of the explanation, though the evidence
here is not straightforward. The first, and most obvious point is that the rate
of GDP growth is hardly the most directly relevant measure of economic performance.
Even though German GDP growth was strong in 2000, it owed much of its strength
to exports. Domestic demand lagged behind. In particular, consumer spending
grew much more slowly in 2000 than in 1999 (by 1.7% compared with 2.5%). The
rate of growth of consumer spending is invariably one of the most powerful indicators
of car demand, and declines, or even very slow rates of growth, of consumer
spending are usually associated with falls in vehicle sales.
“possession of a recently made
car is coming rather lower down the priority list of German consumers than
it did in the past.“
Another striking difference between Germany and the other regions in Western
Europe where car sales were stronger is the much slower rate of employment growth
in Germany. Indeed, in the opening months of 2001, the decline in the unemployment
rate in Germany came to a halt, and even went slightly into reverse. This contrasts
with other countries, most notably France, but also Italy, where employment
growth has continued to be strong, and the car market has held up well.
A further contrast with many other countries in the Eurozone (though not with
France) is that the advent of Euro interest rates meant higher rates in Germany,
while in countries such as Italy or Spain with a higher historic inflation rate,
the level of interest rates will have seemed low, even after the increases put
through by the ECB in the course of 2000.
Even when all of these factors are taken into account, there is still a substantial
part of the fall in demand which is out of line with past relationships. The
idea that consumer preferences are changing is always a last resort, but the
data are consistent with the idea that possession of a recently made car is
coming rather lower down the priority list of German consumers than it did in
What is normal – then or now?
The level of German sales is currently well below, not just the 1999 level,
but also below the average for the post-reunification period as a whole. In
that restricted sense, it is the current level of sales that is abnormally low,
rather than the 1999 level having been abnormally high. However, one might question
whether the level of demand for new cars in Germany was, by international standards,
unusually high during the past, so that what we are seeing is in some sense
a return to normality.
As a rough guide, each million Euros of GDP gives rise to just under 2 new
car sales, across the whole of Western Europe. The number varies widely from
country to country, and is much lower in the Scandinavian countries, where car
acquisition is heavily taxed. In 1999 this number in Germany – the ratio of
car sales to the size of the economy as a whole – was very close to the European
average. Since the decline, it has now dropped well (8%) below the average.
It is true that in the early 1990s this ratio was abnormally high in Germany,
but that differential was steadily eroded in the course of the decade, and,
as we have seen, had been virtually eliminated by 1999. Taking other ways of
“normalising” car sales across countries, for example expressing them in relation
to total consumer spending, provides the same result: on an international comparison,
German car demand was slightly above the European norm in 1999, but is much
more significantly below it now. Thus it is the present that is abnormal, not
just in relation to the past history of German sales, but also in terms of the
West European norms.
How much longer will the weakness last?
At the start of 2001, the boost to disposable income from tax cuts was looked
to as the change that would push German car demand back to normality. So far,
that has not happened, though it is natural to expect there to be some delay
between the tax cuts appearing and an upturn in orders for new cars, let alone
in registrations. Higher fuel prices and other forms of inflation, have, in
any event cancelled much of the boost to spending power due to tax cuts.
The purpose of this note is to describe the nature of the decline, and not
to provide forecasts. But it is not hard to deduce what will be needed to put
an end to the weakness: stronger employment growth, lower inflation, and lower
Oxford, May 30 2001
Auto Industry Statistics – April 2001
– April 2001
|Motor vehicles overall||vf||s||g||-o|
which: foreign makes
up to 6 t gross (excluding Busses)
above 6 t gross (excluding Busses)
Verband der Automobilindustrie (VDA)
To view related research reports, please follow the links