Business cycle and growth policy
Chapter 2: The aim of the full employment
Restriction to the production factor labour
The different dimensions of the employment level
Full employment in the sense of equilibrium
Full employment in the sense of a certain level of employment
The problem of the long-term unemployment
Ascertainable versus necessary economic data
Self-affirmation and self-refutation of prognoses
The problem of a differentiating economic diagnosis
On the problem of the leading indicators
The economic barometer of the council of experts
Growth policy reasons
State political justification
Conflicts between full employment and monetary stability
Conflicts between full employment and income distribution
Restriction to the production factor labour
The demand for full employment
is synonymous with the aim of achieving equilibrium in the labour market. This
equilibrium demand applies actually to all production factors. Here, though, we
confine ourselves to the factor labour.
In the context of the labour factor, additional social aims apply compared
to the other factors of production. In the case of the other production
factors, full employment is especially desirable because without full
employment a waste of scarce resources takes place. A full employment of the
factor labour is additionally desirable as an end in itself. The right to work
of every employee corresponds to the human dignity protected by the basic law.
Thus, with regard to manpower it is not just a question of avoiding wastage
of scarce resources; the right of every single employee for a work of his
choice preferably at free option is striven for his own sake.
The different dimensions of the employment level
The aim of full employment refers initially to the number of employees. The
N = A * a
· N: Total number of demanded
· A: Number of employees
· a: Work time per employee
But the aim of full employment continues and refers to several characteristics.
The aim of full employment relates for one thing to the number of
employees, but also to the work time per employee. A low labour demand could
still be distributed in a way that every worker finds work, only that each
worker has fewer hours of work than he actually wants. The demand for full employment,
however, does not only refer to the aim of finding a workplace for every
worker, but also to the further aim that every worker, only as long as he
desires, fills a complete workplace also. The aim of full employment thus also
relates to the work time per employee.
The work time per employee can for its part be changed in different ways. A
variation of the effective work time per employee is possible.
by changing the legal or
collective working hours,
by changing the average number
of overtime hours,
by increasing or decreasing
short-time work, as well as
by expansion or restriction of
But with regard to the work time per employee it is not just a matter of ensuring
that every worker can achieve as much work time as he wishes; in the foreground
of the political efforts there is often the opposite aim, not of working as
long as possible, but quite the contrary, to reduce the average work time.
In general, the realised welfare is measured not only in terms of average income,
but also in terms of the extent of leisure time. Thus, the unions will not only
fight to raise wages for their members, but also to secure a given wage income
with as few working hours as possible.
The aim of reducing the number of working hours is particularly important
because the individual employee is generally not free at all to decide how much
working hours he has to do. He is usually faced with the alternative of either
accepting a labour supply with a predetermined number of hours that can not be
influenced by the employee, or of abandoning the whole offer.
In addition, it is assumed that the activity of the employee, as opposed to
that of a self-employed person, is largely other-directed, so that an increase
in leisure time, in which the individual can determine by himself how to shape
his own leisure time, is usually regarded as a welfare gain.
Now we must assume, though, that the ideas of the individual employees about
the desired distribution of their time into work time and leisure time differ
greatly. A reduction in working hours is a welfare increase for one person,
because he can achieve greater benefits in his free time than by being able to
achieve a higher income by more work. For another one, though, the same
decrease in the number of working hours is a welfare loss, since he could
achieve a higher benefit increase rather by a higher income, which he could afford
with more working hours, than by more free time.
This conflict, that one and the same change in the working hours influences
the individual welfare for the single employees in different directions, is generally
regulated by the fact that collective or statutory wage rates are binding in
the sense that every employee is entitled for the amount of leisure time
provided by collective agreement, but by way of overtime hours, those employees
who prefer less leisure time, have the opportunity to effectively work more
hours than provided by law.
Due to the different needs structure of individual employees, an optimal distribution
of work time and leisure time would only be guaranteed, if each employee could
decide for himself how he divides his entire time into work time and leisure
time. However, since the prevailing production technique does not allow such
freedom of choice, and presumably if such techniques were possible, then productivity
would fall sharply, the current regulation on the determination of the standard
work time combined with the possibility of working overtime if necessary
represents a second-best solution after all.
Full employment in the sense of equilibrium
It could firstly be spoken of full employment at a match between supply and
demand in the labour market in the sense of total equilibrium. Every person
willing to work will find employment in this case. Every free workplace can
also be filled in this case.
The starting point is the number of all employees in a national economy. A
deduction of the incapacitated is made and of those who are able to work but
unwilling to work. There remains: the number of persons able to work and
willing to work who are employed (the employees) as well as the unemployed.
The degree of employment is here defined as the ratio of employed persons
to all workers: B/A.
B: Total number of employed
A: Total number of employable
workers willing to work
However, there are differences between factual and registered unemployment:
persons registered as unemployed can work on the side. Or out of unconsciousness
and misconceived social prestige, unemployed people do not report unemployment.
Furthermore, due to the protection against dismissal, some employees can not be
dismissed although they can not be employed effectively.
In addition, it may be more advantageous for an entrepreneur to not dismiss
skilled workers despite a lack of demand than to dismiss them and to search and
train new skilled workers again in future needs.
It must also be reckoned with a different development of unemployment on
partial labour markets. The most important reason herefore is an overlay of
cyclical and structural crises.
Other causes also require other therapies. Therefore, it is secondly appropriate
to limit the concept of full employment to the macroeconomic equilibrium. A
possible benchmark for this second term is the ratio of the number of job
vacancies to the number of unemployed persons. At the determination of the
number of vacancies arise difficulties, though. Enterprises are indeed not
obliged to report vacancies to the employment office. What's more, enterprises
partly exert demand for labour force based on suspicion. There may be no acute
need yet, but one wants to get an overview of the market situation and also
conduct interviews, so that when there is a demand for new labour force, one
can immediately hire already tested candidates.
employment in the sense of a certain level of employment
By contrast, William Beveridge had spoken of a breach of the full employment
aim if the unemployment rate exceeded 3%.
The starting point is formed by the following thought: There is a
structurally conditioned long-term residual unemployment which is largely
consistent. Thus, it will not be able to reduce this residual unemployment by
way of employment policies. Then it is appropriate to speak of a need for
employment policy only when the actual unemployment exceeds this mark of 3%.
The attempt to reduce the residual unemployment would be unsuccessful in the
short-term. But then it is also appropriate to refer the concept of full employment
only to workers for whom unemployment can actually be reduced politically.
The residual unemployment can also be fought in the long run, however. But
there are usually other causes that determine the extent of residual unemployment.
Other causes, however, require also other measures. So it still makes sense to
distinguish between the unemployment which is due to lack of demand for goods
and structural unemployment which has to do e.g. with an insufficient education
level of some unemployed persons. While the unemployment which is based on insufficient
private demand for goods may be combated with an expansion of government
expenditure, this measure would have no impact on the structural unemployment,
if e.g. additional government expenditure would be used to develop the
In fact, there has been a change in the ideas about the extent of residual
unemployment in the recent decades. Herbert Giersch, chairman of the Council of
Economic Experts in the 1970s, suggested due to the strong decline in unemployment
that it should not be spoken of full employment until the unemployment rate has
fallen below 0.8%.
Possible causes of this change
according to Giersch are:
the bad weather allowance
regulation of 1957,
an increased mobility,
the active labour market
the role of the guest workers,
a too low education as a cause
Since the 1990s, the unemployment rate has risen again so sharply that one
would now be speaking of full employment if the unemployment rate had fallen to
As causes of this change since the 1990s apply: The reunification brought about
a substantial increase in production costs by way of the solidarity contribution
and therewith deterioration in the competitiveness of the German economy, which
was reflected in an increase in the unemployment rate.
At the same time, globalisation and the involvement of some emerging economies
had worsened international competition, which in turn has led to an increase in
the unemployment rate.
It is therefore entirely justified to refer to the concept of full employment
solely to cyclical induced unemployment and to refer to unemployment which can
be attributed to structural factors as residual unemployment which can not be
combated with traditional economic stimulus programmes. In any case, though, it
must be recognised when it was achieved to reduce the extent of residual unemployment.
In this case, the aim of full employment is also achieved at a different level
problem of the long-term unemployment
It is furthermore necessary to differentiate in the duration of unemployment.
Long-term unemployment according to common understanding is when unemployment
lasts more than a year.
Short-term unemployment as a consequence of welfare-enhancing structural
change is hardly avoidable and is quite tolerable in terms of the consequences.
A welfare increase is only conceivable with a change in technology as well as
in the demand. If the hereby resulting unemployment lasts only for a few
months, then the hereby resulting welfare loss will have to be assessed less
than the welfare loss which would have to be expected if one did not want to
allow any change in demand or any technical improvement.
But long-term unemployment is associated with serious social consequences:
Above all, the reemployment opportunity of the workers who are long-term
unemployed is reduced. An international comparison of long-term unemployment
shows: for a long time, about 14% of the unemployed in the US were long-term
unemployed, while in Europe about 40% and above are long-term unemployed.
Ascertainable versus necessary economic data
As a starting point, we have to realise that different economic phases
require different measures also. Therefore, a situation analysis and prognosis
is necessary, which informs us in which economic phase we are currently. In
times of economic downturn, measures are needed to increase demand for goods
while measures to curb demand are getting necessary in times of overheated
economic upturns, in which demand increases are hardly leading to an increase
in the quantity of goods and therefore fizzle out largely in price increases.
This situation analysis is in practice made more difficult now by the fact
that the economic policy measures take their effect only after a certain time.
We can not assume that e.g. an increase in demand caused by increases in government
expenditure will instantly lead to an increased employment and thus to a
reduced unemployment. Rather, some time will pass until the increased demand on
the goods markets results in an additional demand on the labour markets. Here, one
speaks of time lags.
That these are by no means very short and therefore negligible times has
been shown by Milton Friedman, who pointed out that it would take about one and
a half years between the time when politicians realise that they should stimulate
the economy by employment programmes until these programmes take effect on the
This fact makes it necessary, though, that any economic stimulus programmes
must not only be adopted at the moment when a change in the economic cycle
seems necessary, but just 1 1/2 years earlier.
And in this context arise two fundamental problems. Problem no. 1: in this
case, it is obviously not sufficient, that in the context of a situation
analysis we ask ourselves what the current level of unemployment is and to what
extent this unemployment is contingent on the economic situation. We rather
need a prognosis that provides information about how unemployment will be in 1
1/2 years, namely then, when the measures introduced today will take effect on
the labour market. Problem no.
2: Are the politicians not overcharged, if they should initiate e.g. a policy
of contraction at a time when there is still a considerable unemployment.
Here emerges the fundamental question of whether prognoses within the
framework of the sciences are possible at all. The majority of economists is in
fact convinced that it is possible to make prognoses with scientifically sound
methods. Nevertheless, the prognoses of scientists differ greatly from the
prophecies. Whereas prophecies are formulated namely as unconditioned
statements, the prognoses made by science apply always under certain conditions
The Prophet claims that very specific events will occur at a quite
particular time. The scientist, on the contrary, formulates his statement much
more cautiously. He cannot predict with certainty the event to be prognosticated.
The prognoses are only valid on the condition that very specific conditions are
met. For example, that there are no major upheavals in the political system or
that it has not to be reckoned with bad harvests, either. The prognosis applies
no longer, if one of these conditions has changed.
Furthermore, the scientific prognoses mean that the forecasted event is
only to be expected with a certain probability. Prognoses in the sense that a
certain event will occur with absolute certainty at the presence of certain
conditions, are not possible at all in the context of science. We do not have
absolute knowledge; our knowledge of reality is always more or less imperfect. In principle, we can indeed falsify
universal quantifications (for example, all entrepreneurs maximise their
profits) if we find only one entrepreneur who behaves differently. Thus, in
this case it can no longer be said that all entrepreneurs maximise their
But it is not possible to verify a hypothesis. Only if we could not achieve
to successfully falsify a hypothesis in several empirical studies, we can speak
of a provisionally confirmed theory. But this theory remains always
provisional. We have no possibility of clearly determining whether we have
already gathered all the factors that influence the variable that is to be
explained; we must always reckon with the possibility that certain factors will
be found in the future that will influence the result decisively but have not
been recognized yet, even though they already existed or have not even occurred
Now a few words about the question to which extent the politicians may be
overstrained with initiating dampening measures already 1 ½ years before the expected
overheating, at a time when unemployment is still present to a high extent.
No difficulties would probably be expected if we could assume that almost
all voters had sufficient knowledge of the economy and would act in a completely
rational way. In this case, it would be clear to everyone that the elimination
of cyclical unemployment is only possible if employment policy measures are
initiated in due course, that means at a time when unemployment is still
But in reality, we can not assume that a large section of voters has this
knowledge. The economic connections are so complicated that without economic
knowledge, this insight can not be expected for the majority of the voters.
Under these circumstances, irresponsible populists can exploit this
ignorance of the voters and fight politicians who are already inducing measures
that dampen economic activity at a time when there is unemployment still present.
And in such a situation, it has to be feared that politicians will not take
these measures necessary for success, even if they themselves believe that
these dampening measures are necessary, simply because voters would abandon these
politicians in such a situation. In this situation, the political survival of
the parties is more important than appropriate solutions.
Problems at the determination of the economic situation exist in detail due
the lagging behind of the statistics,
the distinction between inside lags and outside lags,
the distinction between recognition lags, decision lags and realization
We have to assume that there will always pass a certain amount of time
until certain deficiencies are identified. Unemployment statistics, for
example, assume that enterprises which have dismissed an employee report this
to the statistical offices. But we can not expect that such decisions will be
communicated hourly. It has to be clarified also whether these are permanent
decisions. For example, if a casual labourer without a permanent job has not
found work on a particular day, then this does not indicate that the worker's
situation has already changed fundamentally. It would be quite conceivable that
this labourer works only every second or third day on average, so that the
determination of a day off does not say anything definitive at all about
possible changes in the employment of this employee.
The following diagram gives an overview of the different time lags:
07th Self-affirmation and self-refutation of
Prognoses have effects on the data that is to be forecasted. When the predicted
effects occur due to the prognosis, then we speak of the self-affirmation of a
prognosis. The example of an induced price increase is present, if a price
increase is predicted. Here, the demand increases due to the prognosis and the
prices rise along with it.
Whereas we speak of the self-refutation of a prognosis if the predicted
effects do just not occur due to the prognosis. The example of an avoided overcapacity
is present if an overcapacity of production equipment is forecasted. A reduction
of the investment occurs here due to the prognosis and with it the production
Prognoses jeopardise the success of policies when unwanted effects are induced.
For example, the success of the stabilisation policy would be called into
question if the prices would already increase simply on the basis of prognoses.
Prognoses can also be used as an economic policy tool if prognoses induce
desired effects. For
example, the suggestion that overcapacity could be expected in the industry
could already be seen as a means of working towards a reduction of
problem of a differentiating economic diagnosis
The extent of unemployment is - as already mentioned - dependent on different
· frictional and
· structural causes.
Therefore, a theory is required in order to relate unemployment with the individual
Firstly, a distinction between structural and economic causes is necessary:
The long-term average of unemployment corresponds to the structural unemployment;
the fluctuations around this long-term path refer to the economically caused
A distinction between the seasonal and other causes of unemployment can be
made on the basis of experience values: As an example, the above-average
unemployment in the winter months is predominantly seasonal. Since most of the
work in the construction industry is done outdoors, production can not be
maintained, at least when using traditional production methods and at low
Furthermore, the duration of the average job search can be considered as a
benchmark of frictional unemployment. The term frictional unemployment refers
to a worker who remains without a job for a short time at the event of a job
change. If, therefore, the average duration of the job search increases (that
is, the time that elapses from leaving the previous job until starting at the
new job), then the frictional unemployment increases.
The extent of this frictional unemployment is mainly determined by two factors.
On the one hand, frictional unemployment increases when, with the average duration
of job search remaining constant, more workers change jobs or workplaces are
changed more frequently on average.
On the other hand, the extent of frictional unemployment also depends on
the notice periods. For example, if the deadlines by which employees can be dismissed
are extended, then the dismissed workers will be able to start their job search
already at a time when they are still employed in their current job. (Of
course, the cancelled workers must also be given the time to look for new
jobs.) The time that elapses here between leaving the previous job until the
start of the new job, can be significantly reduced in this way, which leads automatically
to a reduction in frictional unemployment.
09th On the
problem of the leading indicators
Historic example of a leading indicator was the Harvard Barometer (W.M.
Persons C. J. Bullock). The Harvard Barometer consisted of three time series:
the speculative curve (dividends and sales of certain shares)
the goods curve (pig iron production, general price index)
the money market curve (discount rate, credit volume etc.)
The precondition for the expediency of a leading indicator is the presence
of clear, dynamic laws:
Wage increases lead e.g. to price increases, but price increases can also induce
wage increases. But we can not assume that the ratio of increases in wage rates
and prices of goods is always the same regardless of whether wage increases
trigger price increases or, conversely, price increases trigger wage increases.
Basically, these relations are based on very different connections, which have
nothing to do with each other.
If increases in wage costs lead to price increases, then it is because entrepreneurs
endeavour to include all costs in the price of goods. After all, they produce
goods in order to make profits in this way. Therefore, they will in any case
try to pass on all cost increases to the price of goods, if possible. But
whether they succeed in any case, also depends on other conditions, for example,
whether the competitors are also facing the same cost increases and are
therefore also passing on the costs to the consumers.
On the other hand, if price increases lead to wage increases, then this connection
is related to the behaviour of the trade unions, their aims of defending the
share of the wage incomes and the real level of the previous incomes.
Furthermore, the strength of the trade unions, which is reflected, among other
things, in the degree of organisation, also determines the extent to which
trade unions succeed in achieving their aims also.
Just because of these different connections, the scepticism prevails today
over the attempt to reliably predict the economic development on the basis of
economic barometer of the council of experts
Here, one confines oneself to the diagnosis and refrains consciously from
making a prognosis.
In a first step, a selection of the economically relevant individual indicators
is made. For example, prices, wages, interest, revenue, unemployed persons, are
selected as indicators.
In a second step follows the normalisation of the economic cycle: The normalisation
is done here via the numbers 1 and 2: The normalised variable values are
increased or decreased by 1, provided that the variables move within the normal
deviations. Conversely, a change of +/- 2 is chosen if the deviations in the
individual variables are greater than normal. Which deviation is considered
normal can be determined by means of the average of the past values.
In a third step, an overall indicator is formed as a moving average of the
On the criticism of this economic indicator the following can be said:
The indicator is not sensitive enough for changes.
It is unclear, for example, what amount of money should be selected: M1, M2
or M3; M1 refers here to the amount of banknotes and coins, M2 includes in addition
the bank deposits created by private banks (time deposits) as well as the
savings accounts, M3 finally considers additionally also other near-money
Furthermore, it is unclear which wage rates should be used, the effective
earnings or the standard wages? For the use of standard wages is indicative
that the individual data can be taken relatively easily from existing collective
agreements. The collection of data is further simplified by the fact that the
number of concluded collective bargaining agreements is relatively small.
The collection of wage data on the basis of the effective earning is in
contrast much more laborious. On the one hand, the data must be obtained from
the large number of individual enterprises, on the other hand, these surveys do
not inform clearly about the extent of the wage shares, which are granted on a
voluntary basis by individual enterprises.
The wage agreements are sometimes so detailed that, for reasons of simplification,
some benefits which are granted on the basis of a collective agreement are not
surveyed in the official tariff statistics. But since all benefits are included
in the statistics on effective earnings, the benefits that are above the
payment agreements are overstated to the extent that the standard wage statistics
do not consider some collectively agreed benefits.
The economic indicator proposed by the council of economic experts becomes
a problem, especially in the case of stagflation: In such a situation quantity
sizes and price sizes cancel each other out; a non-existent stability may
possibly be simulated! Stagflation is known as a situation where general price
increases (inflation) and unemployment (stagnation) occur at the same time.
This danger can be encountered, though, by identifying the price movements
and quantity movements separately and thus preventing that both movements compensate