Part 3: What regulates unemployment?

– What Do the Unemployed Actually Do?

– How Unemployment Regulates Wages

– Booms and Busts

– Wages and Class Struggle

– Union Solidarity

– Arbitration and Wage Indexation

– Capital Accumulation

– Technological Change

– Job Creation and Destruction

Somehow, the size of the pool of unemployed itself must regulate the rates of job creation and destruction. Otherwise the number of unemployed would fluctuate wildly all the time. We shall find out later how the regulation works normally, and why it is not working now. But we already know that unemployment must be some sort of regulator. The larger the pool of unemployment, the more jobs must get created and the less must get destroyed. Otherwise we cannot account for the usual balance eventually reached between these two quite independent rates.

Moreover, we know the mechanism usually tends to reach a balance with only a relatively small pool of unemployed. Therefore the mechanism must continue increasing the rate of job creation and/or reducing the rate of job destruction, as long as there is a certain amount of unemployment. It does not usually stop working with half the workforce still unemployed, just because unemployment is not still increasing.

Finally, we know that whatever this mechanism is, it does not always work. Right now, a small pool of unemployment is not balancing the rates of job creation and destruction. We await each month’s statistics with bated breath to find out whether unemployment has risen or fallen, and we usually find that it has risen. We know that periodically capitalism goes through major upheavals called economic crises, in which a large part of the labour force does get left unemployed for a long period. Our explanation of the balancing mechanism must account for that too.

Whatever the mechanism may turn out to be, we know that as long as it still is not working normally, no amount of artificial “job creation” can prevent the continuing mismatch between normal job creation and destruction from quickly recreating a large pool of unemployed. The only effective remedy for unemployment must be one that gets this balancing mechanism to work again. On this point we can agree with conservative economists. But what is the mechanism, why is it not working normally, and how can it be made to work again?

According to conservative economists the mechanism is simply that increased unemployment tends to pull down wages until it is profitable for capitalists to employ more workers. They conclude that the remedy is push down wages and increase profitability until the unemployment is absorbed.

That sounds quite plausible. If it is true, communists have no reason to deny it. We never claimed that capitalism could permanently maintain full employment without periodically pushing down wages to boost profits. Our answer would simply be that we do not feel like pushing down wages and boosting profits, thank you very much. We would prefer to abolish wages and profits and establish communism.

But a mystery remains as to why the mechanism should not be working, and why the remedy does not seem to work either! To resolve that mystery we shall first have to examine how unemployment regulates wages, and then how wages regulate employment and unemployment. We shall find that there is indeed a close connection between unemployment and wages, and between wages and job creation. But it is not as simple as the conservatives make out, it mainly works in one direction, and it does not work all the time. Most important, we shall find that we can not increase employment simply by pushing down wages. First let’s look at what the unemployed actually do to see how unemployment can regulate wages.

What Do the Unemployed Actually Do?

Under normal circumstances most of the unemployed are not just a stagnant “pool” but an active part of the “stream” moving from one job to another. They form a part of the stream that is temporarily banked up looking for outlets. They are an active part of the stream because they spend their time looking for jobs, not just rotting.

Unemployment is normally a period between jobs rather than a permanent status. When there was “full employment”, half the unemployed at any given time got jobs within four weeks. By 1978 more than a quarter had been waiting for over six months and another quarter for over three months. In a sense, one can measure how “normal” unemployment is, by its average duration, more than by the total numbers involved. It really is not a big problem if the economy is so dynamic that large numbers of people are changing their jobs each year, and they are spending a couple of weeks unemployed between each job. But it is very different when there are actually less people changing jobs than usual, but they are spending a longer time looking.

When unemployment increases slightly, it usually means that people moving from one job to another, or from school to work and so on, have to spend a longer time looking. But it does not immediately mean that a larger number of people are outside the labour force altogether.

Of course some unemployed workers do end up outside the labour force altogether, and even become permanently unemployable as a result of demoralisation. The larger the pool of unemployment, the larger the section of it that ends up stagnating instead of flowing back into employment, and the more peoples’ lives are ruined in this way.

Increasingly the unemployment we have got is taking on the features of a stagnant pool, rather than a flowing stream. This is compounded by the sharp reduction in normal labour turnover as people are reluctant to leave their old jobs unless they have new ones lined up. This stagnant unemployment is a different thing altogether from the “normal” unemployment that somehow regulates the rates of job creation and destruction. Nevertheless, we must first understand how “normal” unemployment does regulate these rates, before we can understand why the new unemployment does not.

The important thing about normal unemployment is that a larger part of the labour force is spending more time looking for work, and not that a section of the population has ceased to be part of the labour force. Hence the concept that the unemployed form a “reserve army of labour” that plays an active role in capitalist production, just as reserve armies play a vital role, and are not simply “inactive” in military battles. Some soldiers are in battle, and others are available to be deployed where required. Some workers work, and others are available to work where they are required. Both those in active service and those in reserve are necessary for things to go smoothly.

An important difference is that reserve armies of soldiers are deployed where their officers decide they are needed. With conscious military planning, reserves can be kept to a minimum and troops transferred directly from one front to another as required. Unemployed workers have no officers and are expected to find their own jobs. (Although there is now a fair bit of “manpower planning” and so forth).

The economic function of the unemployed is to look for work. Those that do not are no longer “unemployed”, but are “not in the labour force”. Those that do will normally find a job eventually. Their place in the unemployment pool may then be taken by someone else looking for employment. How long it takes, and what proportion miss out entirely, depends on the level of unemployment. But the unemployed individuals economic function does not change, their basic situation does not depend on the level of unemployment.

It is important to realise this when attempting to organise the unemployed. One reason they are very difficult to organise is that even now, most of them are not permanently unemployed – and the ones with enough initiative to get organised are also likely to get jobs quicker than average. On the other hand those that do become permanently unemployed can end up getting demoralised and dropping out of the labour force so they are no longer “unemployed” either, and are pretty hard to get involved in anything.

Let’s face it, unless things are really desperate, an individual unemployed worker can get more immediate benefit out of looking harder for a job than out of agitating against the government. The harder you look, the more chance you have of eventually getting to the front of the queue leading back into employment.

How Unemployment Regulates Wages

By looking for work, the unemployed play a vital role in the labour market. Their number determines the ease with which employers can recruit labour for expansion or replacement. That recruitment is going on all the time, even when there is a net reduction in the total number of jobs. There are always vacancies as well as people unemployed (and isolated examples of unfilled vacancies are always pointed to even though there are many times as many people looking for work as there are jobs available). The proportion of unemployed workers to job vacancies determines the average speed with which vacancies can be filled, just as it determines the average length of unemployment.

If vacancies cannot be filled fast enough any other way, then employers will bid up the price of labour by competing with each other to fill their vacancies. As in any other commodity market, this will continue until the supply of labour increases to fill the vacancies, or until the demand for labour has fallen (more likely, since the size of the labour force is relatively inflexible). The demand for labour will fall when the price has been bid up high enough, because investments that would have required more labour will cease to be profitable at the higher wage rate. So less jobs will be created and more will be destroyed. This includes of course the accelerated shift to less labour intensive production techniques.

We will examine the details shortly, but the important point to note is that unemployment only regulates wages in one direction. In fact unemployment only regulates wages when there hardly is any! As soon as there is enough unemployment to avoid a “wages explosion”, additional amounts will not significantly increase the ease with which employers can fill their vacancies.

There is no reason to believe that increased unemployment will cause employers to bid less for labour, or will cause unions to accept less. It may be that with really massive unemployment, union solidarity will be broken down. It may also be that the same slack demand for labour that has created unemployment will also make it unprofitable for employers to bid as much for labour as before. But these are both entirely separate questions. All we know for sure about unemployment as a regulator is that lack of unemployment will drive wages up and that will in turn force employment down.

This one-sided regulation is quite sufficient to explain the observed fact of a normal balance between job creation and destruction with very little unemployment. As long as markets are expanding and there is a tendency for the demand for labour to increase, that tendency will be checked by the size of the available labour force, but will permit full employment and real wages rising together with productivity.

This leaves open the question of whether other factors can also push unemployment and wages up or down and whether unemployment can coexist with high and low wages. That is as it should be, since we know that something other than the normal mechanism must account for the abnormal situation of high unemployment.

By way of contrast, the usual explanation that low wages will increase demand for labour, and high unemployment will push wages back down, explains too much. This would imply that any unemployment will correct itself, when it manifestly does not.

The usual explanation also implies that we should never expect to find increasing demand for labour alongside increasing relative wages. Yet that is exactly what has been happening with increased female labour force participation alongside equal pay.

Finally, the mechanism we have described gives no reason to believe that lowering real wages will automatically produce an increased demand for labour. All it says is that excess demand for labour (more than is available), will cause wages to go up. It does not follow that reduced wages would cause demand for labour to go up. That is also as it should be. Despite all predictions from the conservative camp, unemployment has continued rising while real wages have continued falling.

Since our unemployment regulator only explains one side of wages and unemployment, we need to look elsewhere to find what causes the abnormal movements in a crisis.

First, let us look at what happens before a crisis, namely a boom.

Booms and Busts

In a boom, real wages can even increase faster than productivity, so that the share of wages compared to profits will rise and the rate of exploitation will fall. This actually happened in the 1974 “wages explosion”. That was certainly a “boom” even though there was considerable unemployment at the time. Conversely, when demand is slack, unions have little bargaining power and the share of wages, or even the absolute level of real wages, will fall. That is happening right now.

When the economy is booming there is a general tendency for firms to increase their demand for labour power, raw materials and other inputs, in order to meet the demand for their output. This drains the pool of unemployed, reduces warehouse stocks, increases plant capacity utilisation and drives up wages and other prices. The increased demand for inputs and increased consumer spending can itself feed the boom to a certain extent, by further increasing demand.

But when the price of labour and other inputs is rising faster than the price of firms outputs, profit margins are reduced. There is then a general tendency for firms to cut back their investment and expansion, or even undertake contraction. This reduces the demand for labour and other inputs, and eventually recreates a pool of unemployed as well as stockpiles of other commodities and excess production capacities. Prices and wages are then forced back down (relatively).

These movements occur in individual sectors, but also in the economy as a whole. One firm’s inputs are another firm’s outputs and changes in demand act across the board. The balance between production, consumption and investment depends on movements in wages, prices and the rate of profit. This “balance” is always dynamic since it is precisely the imbalances that bring into play the factors for restoring a balance. Hence there is an unending succession of booms and busts in the economy as a whole and in particular sectors of it.

A problem with the above description of booms and busts is that it seems to describe a self-regulating mechanism that would automatically correct for unemployment or labour shortages by moving wage rates, or the capital intensity of investment, in the appropriate direction. So one would expect things to never get all that far out of balance. Indeed capitalism does work like that, a lot of the time, and normal “fluctuations” in the economy can be adapted to quite smoothly. But there must be more to it than that, when we have a “crisis”. Before going into that, we will look at wages, and then look at the normal balancing mechanism in more detail.

Wages and Class Struggle

Conservative economists assume that left to itself, capitalism always works smoothly, and when it does not, they therefore argue that there must be some institutional factor which is preventing prices from clearing market – for example, unions preventing wages from adjusting to the level of unemployment. Hence the calls for “wage restraint” and polemics against the unions.

To some extent this is hypocritical. Conservative economists are generally well aware that the level of wages is determined by the demand for labour, and not vice versa. They could not really believe that wages are greatly influenced by the effect of polemics against unions. They know that the only effective way to bring down wages is through reduced demand for labour, and that means increased unemployment. So it is quite illusory to talk of unemployment and “wage restraint” as alternatives. They go together.

Even though union leaderships might be very willing to go along with “wage restraint”, the employers themselves will bid up the price of labour power if there is not enough unemployment to hold it down. The propaganda for bringing down wages is really propaganda for accepting mass unemployment.

References here and elsewhere to “unemployment” holding down wages are not meant to imply that competition from the unemployed is the restraining factor. While union solidarity remains effective, there is little such competition. It would be more accurate to say “slack demand for labour” holds down wages. But generally (although not always), slack demand for labour is closely associated with unemployment. So the shorthand reference to”unemployment” is near enough.

Illusions about what determines wages are often spread from the labour movement, and especially its left wing, who sometimes picture the level of wages and conditions as primarily determined by the outcome of sharp class struggles on the shop floor.

This is certainly true to a greater extent than for other commodities. Because of the social elements in wages determination, worker militancy can effect wages more than farmer militancy can effect the price of wheat or supermarket rapaciousness can effect the price of groceries. A militant union can secure more for its members than a weak one and a militant workforce can enjoy a higher standard of living than a more servile one in a country with a comparable level of economic development.

There is an element of real bargaining, and extra-economic factors can also influence the outcome – for example fascist governments that suppress unions, or the threat of revolution. Even so, on a world scale it is clear that the level of wages corresponds very directly to the level of economic development in various countries.

Union Solidarity

The main variable in wage determination is the degree of unionisation and solidarity among the workers. If they are solid, they can get the full value of their labour power – its monopoly price. If they are not solid, they can be forced to accept anything below that – right down to minimum physical subsistence level. Unionisation has been and remains enormously important in raising workers above physical subsistence level and securing the value of their labour power. Smashing unions is still a goal for employers to force workers wages below their value and extract surplus profits.

But once unionisation is well established, unions cannot secure any more than the value of labour power. Like any other monopoly, they cannot charge what they feel like, but only “what the traffic will bear”. In this case “the traffic” is what employers will bid to secure extra labour.

In particular, the main effect of the “level of class struggle” is in determining the overall level of labour conditions for the whole nation. It has much less effect on wages in any particular industry or workplace. Even at a national level, class struggle probably has a greater impact on normal working hours and work intensity and on “social conditions” generally, than on actual wage rates.

Since there is free movement of labour between occupations and industries, the level of wages and conditions in any industry is influenced far more by the overall state of the labour market, than by the level of militancy in the particular industry. Workers in low paying industries will look for jobs in high paying ones, producing a labour shortage in the low paid industry which can only be eliminated by offering higher wages.

The bargaining position of a union also depends more on the demand for its members labour than on the dedication of its leadership. We are talking about variations a few percentage points above and below the wage rate determined by “market forces”. Failure to fight could halve wages compared to their “market” rate. But fighting harder could not double them above the market rate, because the market rate is not some arbitrary figure, but the maximum employers can be made to pay before their investment would be diverted elsewhere. Provided a union does fight, it will get more or less the market rate.

There is a parallel with land rent. Landlords can surrender part of their rent, or have it taken from them in taxes. But they cannot compel capitalists to pay a rent that will leave them with less than the average rate of profit on the capital they invest in the land. The capitalists just would not invest in that land.

Real wages have doubled in Australia since World War II, yet it is not a fundamentally different kind of society. They doubled because of economic development, not because of a sudden upsurge in militancy. Indeed the value of labour power has probably not changed very much. The increase in real wages has resulted directly from the relative cheapening of consumer goods, due to increased productivity.

Arbitration and Wage Indexation

The Australian Arbitration system provides elaborate rituals according to which wage rates are supposed to be determined by impartial judges on the basis of principles of equity. Token 24 hour strikes are an important part of those rituals and feed the illusion that wages are determined by some combination of industrial strength and skill in advocacy.

But arbitration is simply an attempt to measure the bargaining strength of the two sides, without them having to actually waste energy to prove that strength by fighting it out each time there may have been some change. The factors investigated in Arbitration Commission hearings, include the “state of the economy”, “productivity”, “capacity to pay”, “cost of living” and especially “work value” and “relativities”. These are precisely the factors that effect the market determined wage rates. The token strikes are a part of that measurement process rather than a form of real class struggle.

The Commission is trying to determine what the market wage rates objectively are. It does not “set” them. When the Commission guesses wrong, it is soon proved wrong by industrial trouble and/or over award payments and/or sectoral labour shortages or unemployment. Adjustments in the “awards” are then required.

The farce of “wage indexation” is a good illustration. When the Commission really did try to “set” wages according to uniform “guidelines”, it failed miserably. Unions and employers, and finally even the government urged it to allow wages to reflect market forces.

There is no reason to believe that the overall level of wages has been kept either artificially high or artificially low by the Arbitration Commission. The Commission itself is well aware that its centrality in the wage fixing process depends critically on how well it estimates actual labour market conditions. It has as much power to “set” wages as the Prices Justification Tribunal had to “set” (or even “justify”) prices, and less power than the Reserve Bank and the Treasury have to “set” interest rates. These institutions can help smooth things out in their respective markets, and they can stuff things up. But they cannot change the overall direction of market movements.

“Rigidities” have allegedly been introduced into the Australian wage structure by the Commission’s fixed “relativities” between occupations and skills. But this has not prevented wages moving in response to changing demands for labour, nor has it prevented labour moving in response to changing demand. It has simply ensured that the wage movements are slowed down and take the form of over award payments rather than awards. Less flexible “relativities” have encouraged more “manpower planning” to cope with shortages and surpluses of particular occupations, instead of the clumsier process of a change in relative wages having to indirectly attract labour from one occupation to another. Likewise, any “rigidity” in overall wage levels could only produce a time-lag in the effect of underlying market movements.

Leaving aside the hypocrisy, conservative economists do believe that bringing down real wages is an essential part of any program for economic recovery. They have masses of quite genuine statistics to prove that wages have increased more than productivity, the share of wages in Gross National Product has increased and so on. They rightly conclude that there is a “real wage overhang” keeping the economy out of balance, even though the purchasing power of wages may be declining.

Therefore, they see unemployment as necessary to bring down real wages, although they prefer not to emphasise that aspect, but just talk about “wage restraint”. But if more unemployment will bring down wages, why hasn’t it? Why is any “program” for economic recovery necessary at all?

The weak point in conservative arguments is that they do not explain what has changed. It is not good enough to just point out that there is a “real wage overhang” since the “wages explosion”. Why is there, and why have “market forces” not corrected it? Before answering that, we need to look at the normal adjustment mechanism in some detail.

Capital Accumulation

Capitalist production is always a process of production for the purpose of accumulating more capital. One part of profits is spent unproductively by capitalists, maintaining themselves and their retainers “in the manner to which they have become accustomed”. That can be fantastically expensive if you look at the lifestyles of Jackie Onassis and the like. But is a very small proportion of total profits, because there are very few really wealthy capitalists.

The more important part of profits is accumulated as new capital. This does not mean it goes into their pockets, or is hoarded into a pile of gold. It is invested in expanding the wealth and power of the individual capitalist, and incidentally developing the productive forces of humanity.

Capital investment means buying more labour power and raw materials to produce more goods, to be sold for more profits (some of which will allow the capitalist to “become accustomed” to an even more lavish lifestyle, and most of which will be invested to expand further). It is a process of continually expanded reproduction. If there was a fixed supply of labour and a fixed technology, this expanded reproduction would become impossible. The new capital would be trying to recruit workers already employed by the old capital, and it would have no market for its products. Only simple reproduction would be possible, with no net investment.

Even if we allow for increasing supplies of labour, a fixed technology would still only permit expanded reproduction at exactly the rate of labour force growth, with no increase in capital or output per worker.

In fact some “models” of the process of capital accumulation are based on assumptions like that. Naturally, they have not been able to explain very much about real economic growth, which always involves new technology with an increasing social division of labour and more capital and more output per worker.

Real capitalism is always expanding. Hence imperialism. Capital can expand intensively or extensively. It can expand extensively by employing more workers, even with the same technology and the same capital per worker. That is important in the third world where there are still reservoirs of peasants who are not employed as wage workers, and it has been important in pulling women out of the household and into wage labour. It also absorbs population increases.

But capitalism also expands intensively, by investing more capital per worker. This implies new technology and a development of the productive forces, and has made capitalism a far more dynamic and progressive social formation than previous ones which went on reproducing themselves without constantly revolutionising the technique of production.

The increasing organic composition of capital implies a falling rate of profit. Here is not the place for a detailed discussion of that, but it is worth mentioning that the difference in internal rates of profit between more developed and less developed countries is equalised by imperialist capital export and import.

A fuller treatment of capital accumulation should examine it internationally. This is very necessary to combat the narrow nationalist outlook so common in the Australian “left”. Suffice it to say that the unemployment we are suffering in Australia is clearly part of a worldwide problem and will require a worldwide solution. Our industries are not just “foreign owned”. They are part of an integrated world capitalist economy. We should think big.

Technological Change

At any given time, there is always a range of known techniques that can be used for production. This range is also always being extended by the discovery of new techniques, which usually involve the use of new intermediate products and hence an increasing social division of labour. But even without new inventions, there is a range of different ways of doing things, some of which will be economic while others are not.

At lower wage rates and a higher average rate of profit, a given labour intensive technique may be cheaper than a capital intensive technique for doing the same job, even though the capital intensive technique is more productive.

For example, third world countries with little capital invested in roads and so on, are forced to use more labour intensive techniques, even though the more advanced methods used in wealthier countries are already known. It can actually be cheaper to use donkeys for transport until capital is available for investment in building roads and truck plants, producing trucks, training drivers and so on. That capital will not be available until the rate of profit on that kind of investment is higher than on alternatives. As more capital is accumulated, the alternatives with higher rates of profit become saturated, the rate of profit goes down, wages go up, and eventually it becomes cheaper to use a truck. It was always more productive to do so.

One day it may be cheaper to use aircraft for regular inter-city transport. We already know how to, but truck drivers’ wages are not high enough, and the rate of profit is not low enough to justify the massive capital investment required.

Increasing returns to scale often dictate a change in technique when a market has reached a certain size. A road will then be replaced by a railway for example. The railway has greater productivity, but the capital investment required is not economic at low volumes of traffic.

Often increasing returns to scale are associated with a greater social division of labour. Special functions are split away from general purpose establishments and achieve a higher productivity while handling the greater volume. A special repair shop will only become economic with a certain level of repairs. Designs will be produced in-house until their volume permits a specialised design firm to do the job more efficiently.

An increase in the scale of operations as more capital is invested and markets expand, may not be regarded as a change in technique. But there will almost always be changes associated with it, like those mentioned above. In most industries the days are long gone when expansion simply meant that more establishments would be set up using essentially the same techniques.

This increasing social division of labour implies more interconnection between different sections of the economy. Each is producing for all, and all for each. It does not imply greater occupational specialisation. On the contrary it requires greater flexibility in the labour force as they change repeatedly from one job to another.

New capital can be invested in the same old techniques to employ more workers using the same old sort of plant to turn the same old raw materials into more of the same old products. But this is only possible if more workers are available. It always creates jobs, provided there is a market for more of the old product. But it creates no new market and assumes that for some reason the market for the old product has increased.

Otherwise new capital can only be invested in more productive techniques that allow fewer workers (usually using more fixed plant and machinery) to turn more raw materials and intermediate products into more products per worker. This destroys some of the old jobs and creates more or less new ones according to whether the output is increased faster or slower than the labour productivity is increased. That depends on how fast the market expands, which depends partly on how much the new techniques cheapen the product (relatively).

New capital intensive investment is always expanding the market, since it does relatively cheapen the product (or the old technique would continue in use), and since it creates a demand for the additional new plant and intermediate products required by the new technique. If there was no technological progress, capitalism would in fact reach the state of stagnation implied by most economic models, since there would be no expanding market for expanded reproduction.

If output is expanding slower than productivity, the result of capital intensive investment will be less workers employed in that sector of industry. If that is happening overall, the result will be an increasing pool of unemployment since more jobs are being destroyed than created. If output is expanding faster than productivity, then labour intensive investments must be expanding employment.

Job Creation and Destruction

Now we can see how a small pool of unemployment normally maintains a balance between job creation and destruction. As profits are continually being invested to become new capital, old jobs are continually being destroyed and new jobs are continually being created. The balance depends on the relative profitability of capital intensive and labour intensive production techniques in new investment.

As long as wages keep increasing at exactly the right rate to keep on gradually tipping the balance towards capital intensive techniques, investment can continue, with increasing capital per worker, even though the labour force is not growing as fast as capital is being invested (provided there is a market for the products).

Otherwise, if wages fail to grow fast enough, the existing techniques would continue being used by new investments, and this will absorb the pool of unemployed until competition for labour drives up wages and restores a balance.

If wages grow too fast, due to labour shortages, there will be a tendency to switch more rapidly to capital intensive techniques which reduce the demand for labour. Slack demand may then force wages down, but even if it does not, there will be no renewed upward pressure until the labour market has again been tightened by further accumulation.

The point is that in “normal” balance, the demand for labour is directly regulating wages so that demand equals supply. It follows that there can be no such thing as “too many workers” or “too few jobs”. The number of jobs will adapt to the number of available workers as capital accumulates. Likewise, wages will not be “too high” or “too low”. They are determined by the demand for labour. It seems then that “everything is for the best, in the best of all possible worlds”. Mass unemployment is impossible as the textbooks insist.

But this process of adaptation only applies to new capital investments. The existing capital investments can only adapt to labour shortages and surpluses, or changes in wage rates and the rate of profit, within fairly narrow limits. A steel mill cannot employ very much more or less labour according to wages rates. Its design is more or less fixed. Changes can only effect the design of new steel mills, or extensions to plant capacity, and the decision to expand steel production at all.

Lower wages will not encourage existing steel mills to hire more workers. It will only encourage designers of future steel mills to continue using more obsolete, labour intensive techniques. That will only create more jobs when, and if, an increased demand for steel causes more investment in expansion of steel mills.

Thus if there is some disturbance to the normal relationships, unemployment can increase, regardless of wage rates. Contrary to the conventional wisdom of economists, there is no automatic mechanism that would quickly restore a balance. The automatic mechanism can quickly cope with labour shortages, by choking off new labour intensive investment. But it cannot quickly cope with labour surpluses. The unemployment will be absorbed when, and only when, new investment has created new jobs, and that may take some time.

(Next instalment: Technological unemployment)