Chapter 11: Institutional measures





01st Reduction in working hours increases employment?

02nd The dismissal protection legislation

03rd Profit participation

04th Opening of collective agreements




01st Reduction in working hours increases employment?


The Keynesian employment theory is based on the assumption that unemployment is primarily related to the fact that the capital market is not capable of automatically reducing a surplus of savings by way of lowering interest rates. However, it is always implied that an increase in investment and thus also in the expansion of production capacity is certainly desirable and helps to reduce unemployment.


Sometimes, it is argued, however, that Keynesian employment policies would fail because macroeconomic demand should not be expanded at all. This statement is attributed either to the hypothesis that saturation in the demand for goods has occurred due to general prosperity (saturation hypothesis) or that technological progress inevitable leads to more and more work being replaced by machines (robotisation hypothesis). A solution to the employment problem could therefore not consist simply in expanding the demand for goods.


Both hypotheses are wrong. Still, there did not occur saturation for broad sections of the population. Moreover, technical progress is on average only labour-saving if wages are too high in relation to interest rates.


Nevertheless, particularly the trade unions often demand that unemployment must be fought by reducing the working time of each individual employee and that the number of employees could be increased in this way even without increasing the demand for goods.


The main assumption of this approach is that macroeconomic demand for labour hours remains constant with such a measure. In this way it is demanded that jobs are redistributed justly. The basic formula is assumed as follows:

N = const. = A * a ,


where is valid:


N: Number of demanded working hours,

A: Number of employed workers

a: Number of labour hours per employee


A reduction of working time (a) by x% leads accordingly to an increase in employment (A) by about x%. The graphic below illustrates these correlations. We draw the number of employees (A) on the abscissa and the average number of labour hours per employee (a) on the ordinate.


Since we have assumed that a reduction in the number of labour hours has no influence on the total number of labour hours in a national economy (a*A), the correlations between the number of employees (A) and the number of labour hours (a) can be illustrated in an isoelastic curve (red line). It is therefore assumed that entrepreneurs demand a certain number of labour hours and that they do not care how many employees perform this total number of hours.


For example, a total demand for 1 million labour hours could be ensured by 100,000 employees working 10 hours each (per time unit) or 200,000 employees working respectively only 5 hours per day.


Necessarily, a negative change in a (-da) corresponds to a positive change in A (+dA).



There are various plans now for the realisation of an increase in employment by means of a reduction in working hours:


·        Reduction of the weekly working time; the collectively agreed working time is reduced. If, for example, employees have previously worked 40 hours a week in accordance with the collective agreement in force, a new collective agreement would limit working time to 35 hours.


·        Reduction of overtime; overtime contingents are defined which may not be exceeded. Since many employees work more hours per time unit than stipulated in the collective agreement, it is possible to reduce the effective working time by overtime reduction without changing the collectively agreed working time. Often it is also intended that overtime is still allowed, but that it must be used up by holidays within a specified period.


·        Earlier retirement of the labour force. The age from which employees can retire from the labour force and therefore receive pensions will be reduced. In this case, for example, employees retire at the age of 64 instead of 65.



The following criticism is directed against these plans:


Firstly, macroeconomic demand for working hours is not constant in reality. Certain tasks, especially management tasks, cannot be divided arbitrarily.


Secondly, the inflexibility of labour demand is a consequence of institutional regulations and is not predefined. In itself, the demand for labour depends on the level of incurred costs. A different distribution of tasks among individual employees requires changes in technology, with the result that costs rise. If, for example, a secretary's work is divided into two half-time jobs and the working schedules of both part-time jobs do overlap, then additional workplaces must be created.


Thirdly, the freedom of the individual employee is severely restricted here. It is actually desirable that, within the framework of technical possibilities, individuals should be able to decide for themselves as needed to what extent they wish to extend their working hours by working overtime. If the possibilities of overtime are limited, the employee can no longer realise his individual plans with regard to leisure time and working time.


Fourthly, the unemployment problem is being solved by a worsening of the pension crisis. If these plans provide for early retirement, one problem of unemployment will basically be solved at the expense of another problem, the security and financial viability of the pension system.


The statutory pension insurance system is faced with the problem that due to demographic developments fewer and fewer employees paying contributions are attributable to one pensioner. This means that contributions to the pension insurance of an individual employee must cover an ever-increasing proportion of pension rights of the pensioners.


In order to avoid either a significant increase in contribution rates or an extremely low level of pensions, the year in which employees leave the labour market should be increased, actually. With the plans to reduce working hours exactly the opposite is achieved. Employees are leaving the labour market earlier than before.


Fifth, an overtime reduction decreases the adaptability of markets, thus further increasing the risk of unemployment. The possibility of extending effective working time by overtime basically solves two problems. On the one hand, it grants the individual employee the opportunity to work more hours than it is provided in the collective agreement. On the other hand, enterprises can thus avoid temporary imbalances between supply and demand of goods.


If, for example, an enterprise receives additional orders after a long recession, but cannot be sure yet that these orders will continue in the future, thus a sustained economic upswing had already been initiated, they can cover these orders by working overtime.


As long as entrepreneurs cannot be sure that demand has risen in the long term, they will not hire new labour force, since due to a restrictive dismissal legislation these employees could not be made redundant in the short term if this additional labour force will no longer be needed due to a lack of demand for goods. In these cases, entrepreneurs will forgo these orders if they are deprived of the possibility of overtime work. But just in this way, the danger of additional unemployment is even greater than usual.


This employment formula encounters difficulties especially in small enterprises. Newly recruited workers should be able to carry out a wide range of activities, since not entire workplaces are cancelled. Let us take as an example a small enterprise with a total of 10 employees. By reducing the weekly working time of employees by 10%, a new worker shall be hired. However, it is unlikely that this aim can be achieved in this enterprise, since each employee had a different task to perform and the new employee should therefore be able to perform respectively 10% of the work of all the other 10 employees.


A satisfactory solution can only be achieved in an enterprise which, with a 10% reduction in weekly working time, has at least 10 jobs for each individual occupational group. If, for example, the enterprise had previously employed 10 foreign language secretaries, an additional position could be created for foreign language secretaries in which simply each of these 10 employees would hand over 10 percent of his tasks.


A success of this reduction in working hours can furthermore only be expected if there is no complete wage compensation, i.e. if the following demand:

L = l * a = constant


is forgone.


We must consider that the demand for labour depends on the level of costs. The costs consist mainly of labour costs. The higher the unit cost of labour, the lower the number of labour hours in demand. But full wage compensation means that to the extent that working hours are shortened, the wage rate increases, because only in this way the wage income (l*A) remains unchanged. Thus, the demand for labour will only remain unchanged if wage compensation would be completely foregone. The higher the wage compensation is, i.e. the less the employees must accept a reduction in their wage income, the greater may be the resulting decline in labour demand.


The costs increase generally also because the plans to reduce working time create additional capital requirement. This increases costs and, with them, unemployment rises again. However, the extent to which costs increase and labour demand decreases depends on labour demand elasticity. We want to assume that the elasticity of labour demand was > 1. As the graphic below shows, this elasticity value results automatically if we assume a Cobb-Douglas production function.




The graphic below shows how a reduction in working hours with full wage compensation and the assumption of a Cobb-Douglas production function affects employment. We draw these connections into a four-quadrant system. In the northeast quadrant we see the employment function, which shows how wage changes affect the demand for labour. The northwest quadrant shows the correlations between wage rate and working time. If we assume full wage compensation and thus a constant wage income, then wage income can be displayed as an isoelastic curve.


At the beginning, so be assumed, the working time was a1. At a wage rate of l1 there was a demand for labour of N1. Now the working time was reduced to a2. In order to achieve full wage compensation, the wage rate must rise to l2. At this wage rate, however, the demand for working hours drops to N2. In the southwest quadrant we can read from the tangent angle (A) how the demand for employees has changed. The graphic shows that demand for labour has also fallen from A1 to A2 as a result of the working time reduction.


Of course, the exact change in the number of workers depends on the parameters of the production function. Not in every case is to be feared that labour demand will even decline, even though this measure was introduced actually to reduce unemployment. In any case, however, it can be shown on the basis of these correlations that the increase in the employment of workers does by no means reach the level of the reduction in working hours. In practice, we must assume that a 10 percent reduction in working hours will lead to an increase in the number of employees of significantly less than 10 percent.




02nd The dismissal protection legislation


In Germany, dismissal protection is determined by law, collective agreements and jurisdiction. The aim is an avoidance of arbitrary dismissals. The legal protection against dismissal is justified by the fact that in this way a natural monopoly position of the employer or at least its effects shall be avoided.


Does this labour market situation still apply today? In the past there was a natural monopoly position of the entrepreneur due to a low regional mobility of the workers. Today, the regional mobility of most employees is significantly higher. Nevertheless, the natural monopoly of the employer remains, often due to an information monopoly of the employer. However, a more effective protection against unemployment would also be given in the case of competition. If there was genuine competition between enterprises on the labour markets, individual employees would be able to offer their labour to another entrepreneur in the event of dissatisfaction.


Even if dismissal protection is able to largely prevent arbitrary dismissals of employees, it will not increase the number of workplaces! In contrast, dismissal protection increases costs and thus reduces the demand for labour and in this way even increases unemployment.


On the one hand, it must be expected that, due to a decline in the demand for goods, entrepreneurs will no longer need a part of the employees previously employed for production. But nevertheless, they will not be able to dismiss these employees because of the dismissal protection.


In addition, a poor sales situation can force an enterprise to rationalise, in this way reduce the cost of production and actually for this reason lay off labour force. But this rationalisation, which is necessary in order to improve the competitive situation, can fail because employees must not be made redundant initially due to the dismissal protection.


Here it may even be the case that an enterprise loses its competitiveness precisely because it must largely forego rationalisation and must just therefore file for bankruptcy. In this case, unemployment has even increased finally, despite or precisely because of the dismissal protection legislation.


In this context, conflicts arise with the interests of the unemployed, especially the long-term unemployed. Precisely when the entrepreneur is unable to dismiss existing labour force due to dismissal protection, he may also hire new labour force to a lesser extent, so that the already unemployed have greater difficulty in obtaining new employment. Dismissal protection may increase the probability of already employed workers not to become unemployed, but at the same time the same legislation reduces the chances of already unemployed people to find a new job.


But this distribution of job opportunities is undesirable for reasons of both efficiency and justice. Enterprises can no longer choose freely to employ the respectively most productive labour force. Once a worker is employed, he has much better chances on that workplace than another unemployed worker, regardless of his ability, even if the unemployed person actually has the higher qualification.


But since just the unemployed are much more in need of protection than all other employees who are already employed, this regulation also violates the principle of justice on a massive scale.



03rd Profit participation


Proposals for employee profit sharing usually aim to change the distribution of assets in favour of employees. Martin Weitzman, by contrast, proposed reducing unemployment by introducing profit sharing. The plan developed by Weitzman attempted to achieve this aim by shifting the forms of remuneration to performance-related and earnings-based wages. In this context, Weitzman developed the following thesis:


The demand for labour is ceteris paribus higher with an earnings-based wage than with fixed wages. Weitzman saw the reason for this in the fact that a profit share would lead to a lower entrepreneurial risk. At the same time, profit sharing would accelerate the adjustment to data changes, so that the adjustment losses associated with data changes could be reduced.


The following graphic illustrates the way profit sharing works. We draw the produced quantity (X) on the abscissa and the achieved price (p) for these goods on the ordinate. The red line shows the supply curve, which shows the quantity of goods which the entrepreneurs are willing to offer at alternative prices. Whereas the blue line indicates the demand of households for alternative prices.


In addition to these two curves, we draw a curve in our diagram that shows the course of the profit sum for alternative quantities of goods. A positive profit is not achieved until a certain quantity of goods, since certain fixed costs incur even if the enterprise would not produce anything at all. The profit initially increases as the production volume increases, since the difference between the marginal benefit for consumers and the marginal costs is initially relatively high yet.


But as the marginal benefit of the consumer decreases with an increasing quantity of goods and at the same time the marginal costs of the enterprise rise, the profit increases that are achieved by an expansion of production also become smaller and smaller. With a certain quantity of goods - exactly where the supply curve and the demand curve intersect - the profit reaches its maximum and then decreases again with further expansion of the production quantity.




Before the introduction of profit sharing, the balance between supply and demand was achieved for a quantity of goods X0. It is now assumed that entrepreneurs will pay part of their employees' wages by means of profit sharing. This means two things. On the one hand, the employees share in the profits of the enterprises, on the other hand, their fixed wage decreases, whereby the total income of the employees (wage sum plus profit sharing) is not changed in comparison to before. In our graphic, we record the employees' profit share by subtracting the respective profit share accruing to the employees from the gross profit.


In this way, we obtain a second profit curve that reports the net profits remaining to the entrepreneurs. This follows the curve of the gross profits. Where the gross profit is zero, the net profit is also zero, and for the quantity of goods where the gross profit reaches its maximum, the net profit reaches its highest amount, too. Only that just the net profit makes up only a certain part of the gross profit always.


The conversion in the form of remuneration will automatically lead to a downward shift in the supply curve. We remember: the supply curve is derived from the marginal cost curve; the entrepreneurs extend their supply as long as the marginal costs are lower than the price obtained at the sale. A reduction in fixed wage costs, however, leads to a reduction in marginal costs at each output, the marginal cost curve and thus also the supply curve shift downwards.


In order to better perceive the impact of this change in the form of remuneration of employees, we now additionally want to assume that the demand for the goods declines as a result of the economic situation. This change in the data has the result of a downward shift in the demand curve. Demand has declined at all possible prices. We now want to examine how this decline in demand for goods affects employment at fixed wages on the one hand and on the other hand when part of employees' wages is paid as profit sharing.


Without profit-sharing, the new market equilibrium (the new intersection of the supply curve and the demand curve) would lie at a quantity of goods X1, thus at a lower quantity of goods, due to the decline in demand. But in contrast to this, the new equilibrium point for employee profit sharing is at the quantity X2, which is larger than without profit sharing. Since employment clearly depends on the production volume and therefore a larger production volume also corresponds to a larger employment, the introduction of profit sharing has actually led to an increase in employment in this way.


Employees have a higher income risk with profit share, but a lower employment risk. According to the ranking of employees, job security always ranks prior to income growth. This means that the overall wellbeing of the employees would be met in this way.


With profit sharing, a motivation-based productivity increase is also possible. Since improvements in production are automatically reflected in higher wages, it is in the interest of employees to increase productivity by means of special care and effort. But as productivity increases, so does the ability to provide employees with higher incomes, too.


The introduction of profit sharing in line with Weitzman's ideas could also help to lead out of the current impasse on the labour market. In the first half of the 1990s, wage increases exceeded labour productivity growth, but employment remained constant in the short term, although one might have expected that this expansionary wage policy would either have led to price increases, if the central bank had been ready to finance the income increases by expanding the money supply, or else would have triggered a reduction in employment.


In the mid-1990s wage increases were lower than productivity growth, but employment did not increase in the short term, although it should have been expected that this moderation by trade unions would have led either to a decline in price increases or else to an increase in employment. How can these contradictions be explained?


These developments do not mean at all that the previous theses, according to which wage increases that are above the increase in labour productivity either increase prices or else - depending on the attitude of the central bank - reduce employment, are wrong.


The idea is wrong, however, that these negative effects of an expansive wage policy have an impact immediately after the conclusion of these collective agreements, just as it is wrong to expect that a moderate wage policy of the trade unions in the sense of a limitation of wage demands on the increases in labour productivity must have a beneficial effect immediately afterwards.


We need to realise that the feared or hoped-for effects of the union wage policy will be achieved by making it beneficial for enterprises to switch to a different capital intensity. We have already seen elsewhere that the question as to which capital intensity is targeted by the enterprises can usually be decided only the way they change production technology as part of their investment programmes.


Here, a distinction must be made between a short-term and a long-term view. In the long run, any increase in the wage/interest ratio means that more capital-intensive procedures are worthwhile. But this does not mean at all that an enterprise responds in the short term to these changes by changing its capital intensity immediately. Investments are always used to construct plants which determine the production conditions for several years. Only when a production plant has been in operation for several years and has therefore been largely amortised, it makes sense for an enterprise to change the production plant.


But this also means that changes in wage policy will only have an impact on price levels or employment in the long term, i.e. two years after the change. The moderation in wage demands only pays off for employees in the long term, since employment only increases again when new workplaces are created as a result of new investments.


Precisely because these correlations are quite complex, and therefore by no means immediately recognisable by a layman of economic science, there is the danger that an expansive wage policy can last for a longer period even if the long-term effects also mean a damage to the employees.


Furthermore, for trade unions that have adhered to the principles of a productivity-oriented wage policy and enforced moderate wage demands, it seems that the desired improvement in price levels and employment has apparently failed to occur. Then it is difficult for these trade unions to continue with the moderate wage policy because it appears that it has not proved itself and there is a danger that even with good will and conviction that this policy will be successful in the long run, trade unions will find it difficult to defend this policy before their members because it appears that employees are worse off because they forgo expansionary wage demands without necessity.


Profit sharing constitutes a way out of this impasse. Employment is rising, as at least a medium-term commitment of labour income and thus greater flexibility is achieved. Entrepreneurs can assume that the income of employees will only increase in the medium term if earnings increase, too.



04th Opening of collective agreements


The starting point constitutes the fact that we have a decentralised, but intercorporate collective bargaining practice in Germany. It is indeed allowed that the wage conditions contracted in the collective agreements are supplemented by additional operational agreements between an enterprise and the works council. However, the principle of indispensability applies. According to this principle, the arrangements made in the operational agreements must under no circumstances result worse than those in the collective agreement. As argument in favour of the adherence to intercorporate collective agreements it is argued as follows: it is more efficient and fairer, too, if similar problems are solved in intercorporate agreements.


This principle of indispensability is interpreted very rigorously by the labour courts now. For example, an operational agreement which provides for lower wage rates than those provided for in the collective agreement is regarded as legally void, even if the acceptance of a lower wage would entail that the enterprises commit themselves not to make any reductions in employment, or at least to make fewer reductions than previously planned.


Now it has been repeatedly stated in surveys that employees mostly give higher priority to job security than to achieving higher wage rates. In this case, however, an occupational contract that provides for greater job security at a discount on the wages to be paid would clearly increase the welfare of the employees. And in this case, it is not justifiable to declare such contracts as legally prohibited. It would therefore be much better if the labour courts, in their decisions, were to examine less the question of the extent to which individual provisions - such as the setting of wage rates - are undercut by the operational contract, but examine more the question of how the overall welfare of the employees is influenced by these additional agreements.


Of course, one must admit that it is much more difficult to make such assessments of actually subjective variables than to determine whether certain wage rates (or other working conditions) are undercut. Nevertheless, the example of the surveys shows that a higher evaluation of job security prior to the aim of a wage increase can still be objectively proven in concrete individual cases.