Summary:
Michael Roberts in an article in WW says COVID has blown a hole in mainstream economic theories of inflation, he's right, but his analysis and alternative is also wrong.
His main object is to counter the Monetarist theory, but, although he also seeks to oppose the Keynesian theory, he ends up with a version of Keynesian cost-push inflation.
He objects to the Monetarist definition of inflation as a monetary phenomena, but that is also what Marx says in his analysis of money and inflation.
Roberts' analysis might be correct if currency were still in the form of a money commodity like gold, but it isn't, and, the shift to money tokens, as Marx described in A Contribution to The Critique of Political Economy, be they coins, paper notes, (or now electronic digits), changes things, and, in fact, turns it on its head.
The quantity of a money commodity put into circulation is determined by the value of commodities to be circulated, i.e. the average value of all commodities multiplied by the quantity of commodities to be circulated, divided by the value of the money commodity, multiplied by the velocity of circulation, in other words, the equation MV=PT that Roberts objects to. As Marx describes, where a money commodity, such as gold or silver, is used as currency, it has to be the value of commodities that drives this, because, if more value of gold is put in circulation than is required, the value of the currency, say an ounce of gold, falls below the value of gold as a commodity or bullion. Gold as currency would be taken out of circulation, and converted to bullion or used as a commodity, for use in jewellery etc. But, this cannot occur when money tokens replace the money commodity, because they do not possess the value of the gold they represent.
Because of this relation of currency to the value of commodities to be circulated, its apparent what the problem with Roberts' argument is. He says the underlying determinant is changes in the value of commodities. The determinant of values, however, is social productivity. In pre-capitalist modes of production, differences in productivity, in different spheres, bring about variations in these values, and consequently exchange values, i.e. the proportion in which one commodity exchanges for another. The value resolves into c + v + s. If, productivity, in one sphere, remains constant, so that v + s remains constant, whilst a fall in social productivity causes v, i.e. wages, to rise, then the consequence is that s must fall proportionally, leaving c + v + s unchanged. However, under capitalism, commodities exchange at prices of production, not exchange values. The price of production is k, the cost of production (c + v), plus the average profit, p. So, now, if social productivity falls, and wages rise, this will also affect the price of production of the commodity, even though it does not affect its value! This applies to gold, if it were the money commodity too. In other words, under capitalism, the prices of all commodities are determined by the prices of all other commodities.
If we put this in terms of exchange-values, then, a change in social productivity would affect the value of gold in the same way it affects every other commodity. If we assume that the velocity of circulation is 1, let 1 ounce of gold have a value of 10 hours labour. Let this unit of currency be called £1. The total value of commodities to be circulated is 1 million hours, say 100,000 commodities, so each with an average value of 10 hours labour, or £1. The amount of currency to be put into circulation then amounts to MV=PT, and so £1 x 100,000 = £100,000 = 100,000 units x £1. If there is a fall of 10% in social productivity, so that the commodities now require 1.1 million hours to produce, so each, on average, has a value of 11 hours labour, how many currency units are then required, and how does this affect prices, and inflation? Well, gold is also a commodity, which is a requirement for it being the money commodity. However, the same fall in social productivity means that 1 ounce of gold also now requires 11 hours of labour to reproduce, i.e. the value of 1 ounce of gold is now 11 hours labour. The 1 ounce of gold, called £1, thereby, also now represents 11 hours labour. Dividing the total value of commodities to be circulated, of 1.1 million hours, by the value of 1 ounce of gold, which is now equal to 11 hours labour, means that 100,000 currency units are required, as before, and the total prices of commodities remain at £100,000. In other words, there would be no inflation, even though the fall in social productivity means that the total value of the commodities being circulated has risen by 10%.
This demonstrates the difference between values (measured directly in labour-time) and exchange values (measured indirectly in quantities of other use values), and part of the error in Roberts' argument is that he fails to make this distinction. If all values rise equally, there is no change in exchange values, including the exchange value of the money commodity to all others.
This assumes, of course, that the 10% reduction in social productivity still allowed the same volume of commodities to be produced and circulated as before. That would require, in fact, that additional labour could be employed to produce this additional value. In other words, it would require workers to work 11 hours during the day rather than 10, or would require 11 workers to be employed, where only 10 were employed before. If this were not possible, then the fall in social productivity would mean that fewer commodities were produced. In that case, with the value of the money commodity itself increased by 10%, fewer currency units would be put into circulation. For example, suppose with a fall in social productivity there is only sufficient social labour-time available to produce 90,000 commodities, each now with an average value 11 hours labour = 990,000 hours value. But, 1 ounce of gold now also has a value of 11 hours, due to the same fall in social productivity. In that case, only 90,000 currency units have to be put into circulation. The name of each currency unit would remain £1, and the average price of a commodity would remain £1. There would be no inflation of prices.
It is quite clear, then, that inflation cannot be the result of changes in values, i.e. from an increase in costs due to a fall in social productivity. It can only result from a change in the relation of the value of commodities to the value of the money or the tokens that represent it. Inflation is a monetary phenomena.
Suppose we swap the 1 ounce of gold, for a paper note. The paper note, now, becomes a proxy for the 1 ounce of gold, or, more correctly, for the amount of social labour-time it represents. If, 1 ounce of gold now represents 11 hours of labour-time, then so must the token that takes its place. In the latter example, the amount of paper notes in circulation must be reduced, as the total value of commodities has fallen to 990,000 hours, whose equivalent form is 90,000 ounces of gold. In the former example, 1 million notes would be in circulation, as equivalent form of the 1.1 million hours of labour time embodied in the commodities to be circulated. In both cases, the average price of the commodities remains £1, so there is no inflation. It is not the increase in the value of commodities that causes inflation, but the change in the relative values of commodities, in respect of the money commodity, or its token. Inflation is a monetary phenomenon.
When precious metals were used as currency, or the money commodity out of which coins were produced, inflation could arise when the value of the money commodity fell relative to the value of other commodities. So, for example, when the Spanish conquered South and Central America, they obtained vast amounts of silver. When it was taken back to Spain, this lower value of silver, used as currency, led to an inflation of prices. When gold takes over the role of money commodity from silver, the same thing occurs with vast amounts of gold pillaged and sent back to Europe, and when large new sources of gold are more easily acquired, from the various gold rushes, this has the same effect of devaluing the money commodity, and, thereby, its relative value (exchange value) to other commodities, causing the prices (exchange-value expressed as a quantity of the money commodity) of those commodities (inflation) to rise.
Inflation arises when money tokens are put into circulation representing a greater value of the money commodity than would otherwise have been required. This is true whether those tokens are comprised of the money commodity, or are comprised of other materials, including paper, or are simply electronic digits, registered in an account on a computer. If a £1 coin comprises 1 ounce of gold, and represents 10 hours of social labour, it is not a matter of whether this coin is debased or not that results in inflation, but merely the quantity of such coins put in circulation, relative to the quantity required. If 100,000 gold coins are required to circulate 100,000 commodities, representing a total value of 1 million hours of labour, then if 110,000 full weight coins are put into circulation, each coin will be depreciated by 10%, even though its value, as determined by the gold content of the coin, will remain 10 hours of labour. It would now require 11 such coins to buy 10 ounces of gold. Consequently, it becomes logical for the owners of such coins to take them out of circulation, to use them as bullion, to melt them down and sell the gold itself for use as jewellery etc. By this means, the number of such coins in circulation would be reduced.
The determinant is not the gold content of the coin, but the relative quantity put in circulation. If the gold content of each coin were to fall to half an ounce due to wear, or clipping, so long as there remained in circulation only the same 100,000 coins, they would continue to function as though they were full weight, because the coins act only as token for the 1 ounce of gold they represent, which, in turn, acts only as representative of a given amount of social labour-time. If half the coins in circulation are half weight, and the rest are full weight, but twice as many coins are put in circulation as required, the value of each coin will still act in the same way. Every coin, full weight or half weight, will be depreciated by 50%. It is depreciated as a coin, as a token of the money commodity, not as a commodity. In such conditions, full weight coins can be removed from circulation. They cease being currency/money, and become commodity. The half weight coins could not, because the value of the gold in them would be no more than the depreciated value of the coin itself, but it is not the diminished quantity of gold in the coin that devalues it, but the excess quantity of currency in circulation. For this reason, base metal coins or paper notes cannot be removed from circulation, when the issue is excessive and converted into commodity. That is why, as Marx describes, the situation becomes turned on its head when these worthless tokens take the place of precious metals as currency.
Once these worthless tokens take the place of precious metals as currency it is no longer the value of commodities to be circulated that determines the amount of currency to be put in circulation, but the quantity of currency put in circulation which determines the average prices of the commodities being circulated. The currency tokens become devalued when they are issued in excess, and so the values of commodities, becomes expressed in a higher money price – inflation. Inflation is a monetary phenomena.
Roberts argues that more money has been put into circulation, and yet prices have not risen. Firstly, that fails to take into account the effect on the values of commodities resulting from massive rises in social productivity from the 1980's onwards. If commodity values fall, then prices should fall too, but if they remain the same or rise slightly in money terms, that is an indication that there has been inflation, i.e. excess currency was put into circulation so that the exchange value of the currency relative to commodities has fallen. Secondly, its not true to say that prices have not risen. Roberts himself admits that the prices of assets have risen over that period. Assets are also commodities bought with currency. A look at the rise in the Dow Jones since 1980 shows the massive inflation of asset prices during that period, resulting from the excess amounts of currency put into circulation. Look at the rise in bond and stock markets over the period of the lock-downs, which is a direct result of the massive increases in currency put into circulation over that period, despite the fact that the revenues to those assets were sharply curtailed during that period.
The fact that it was asset prices rather than commodity prices that rose sharply is a reflection of what Marx points out in the Contribution that having put currency into circulation the authorities cannot control where it goes, and so which prices it will inflate. More correctly, in this case, it is an indication that central banks did exert significant control over where the currency went, by directly using the currency to buy financial assets, and encouraging others to follow suit, in search of large capital gains, and so also to drain liquidity from general circulation. The fact, that, during the lockouts, those spheres where consumers were able to spend, saw large rises in inflation, as the excess currency swilled into them, is further indication of that. The fact that, as lockouts end, and money demand increases sharply, as all of this excess currency is mobilised, prices of commodities across the board are rising by large amounts, is further evidence of that reality. Inflation is a monetary phenomena.
Roberts wants to argue against the Keynesian cost-push theory of inflation, but his own argument is a version of cost-push inflation, phrased as an increase in values rather than costs. That is hard to justify, given that, under capitalism, prices are determined by prices of production, not exchange values, and prices of production are a function of cost of production plus average profit. Roberts also makes the same mistake as the Keynesians, by confusing the superficial appearance for the underlying reality. Both are right in seeing that rising costs/values lead to rising prices, but fail to identify the underlying mechanism. If we take a money commodity, then, as described above, its impossible for rising costs/values caused by falling social productivity to cause inflation, because the value of the money commodity would increase in proportion to the value of other commodities. Its only where there is a disproportion that inflation could arise, i.e. the value of the money commodity/token falls relative to the value of the commodities it is to circulate. In a regime of prices of production, determined by costs of production and average profit, the prices of all commodities, including the money commodity are determined by the prices of every other commodity. If social productivity falls, then the prices of inputs for gold producers, for example, including the costs of wages, will rise, so that the cost of production of gold rises. Applying the average profit to this cost of production then gives the price of production of gold, around which its market price revolves.
Roberts adopts the cost of production theory of value of Adam Smith, as against Marx's labour theory of value. Suppose that we have c 1000 + v 1000 + s 1000 = 3000, where these are values measured in labour hours. Now, social productivity falls in Department I by 10%. To reproduce the constant capital, requires 1100 hours. Smith, and Roberts now calculate the value of output as c 1100 + v 1000 + s 1000 = 3100. However, this is impossible unless an additional 100 hours of new labour are undertaken during the year. The value of the consumed constant capital was given as 1000 hours. (If that represented 1000 units, the fall in productivity means that say only 900 are produced, this year, so that the unit value of each is increased, reflecting this rise in the value of the consumed constant capital) As a result of the fall in social productivity, in order to reproduce this consumed constant capital, 1100 hours are required, and in resolving the value of this years output, the increased value of constant capital would have to be accounted for. This is what Marx sets out in Capital III, Chapter 6, and in Theories of Surplus Value, Chapter 22, in relation to The Tie-Up of Capital. In other words, if only the same amount of new labour is undertaken during the year, the consequence is that a portion of the surplus value, is now tied up to reproduce the consumed constant capital, otherwise it is not physically reproduced, and social production would be contracted. As Marx says, it creates the illusion of an actual fall in profit. So, if no additional labour is undertaken, the output of the current year, with a value of 3000, then resolves as follows: c 1100 + v 1000 + s 900. In other words, it appears that the surplus value has fallen by 100, even though, there is no change in the rate of surplus value, i.e. 2000 of new value is created by labour, and only 1000 of that new value is required to reproduce the consumed labour-power. What it really represents is that a portion of total current output now has to go to replace the consumed constant capital, whose value has been increased retrospectively. If production is to continue on the same scale, the same quantity of labour must be employed as before, determined by the technical composition of capital, and so 1000 of the current output must again reproduce the variable-capital, which means that 100 hours of current output that would previously have gone to surplus value, no longer exists, which means that the unproductive consumption of the capitalists, or else the amount they would have allocated to capital accumulation is reduced by this amount. As Marx describes in Theories of Surplus Value, Chapter 22, although this apparent reduction in surplus value is an illusion (it is a conversion of revenue into capital), what it does constitute is a fall in the rate of profit. In other words, even if we disregard the fall of 100 in profit, the rate of profit would fall from 1000/2000 = 50% to 1000/2100 = 47.62%. This is clear when looking at the next year. In the next year, with no further change in social productivity, we would have c 1100 + v 1000 + s 1000 = 3100. In other words, the 1100 hours value of constant capital is now included in the value of current output, and reproduced out of it. The employed labour again produces 2000 hours of new value, and is resolved into 1000 to reproduce variable-capital, leaving 1000 as surplus value. But, this 1000 of surplus value, whilst still constituting a 100% rate of surplus value, and despite the total value of output now having risen to 3100, or indeed because of it, now represents a rate of profit of 47.62%, not the 50% that existed prior to the fall in social productivity.
What the Keynesians and Roberts see is rising costs/values followed by rising prices, but this is the consequence of the currency taking the form of money tokens. Take the economy as a whole, so that we can ignore the question of exchange-values as against prices of production. Now, the value of output resolves into c + v + s. Let us say, this is equal to 1 million hours of labour, represented by 100,000 paper notes each called £1, and representing 10 hours of social labour-time. If social productivity falls/costs rise, then as Marx describes in Capital, the increase in the value of c will be reflected in the value of the output, out of which it is reproduced. The new value created by labour (v + s) will also increase, because more new labour is required. But, this increase in costs means that more labour-time is required to reproduce labour-power, so that v will rise relative to s.
If money took the form of a money commodity, such as gold, then these increased values/costs would also be reflected in the value of gold, so that the price of commodities, i.e. their exchange value measured in gold, would be unchanged, unless there was some disproportionate change in the value of gold compared to all other commodities. If 1 ounce of gold is given the name £1, then the gold price of all commodities would remain unchanged.
But, the currency is not comprised of units of gold put into circulation. It is comprised of tokens, of generally worthless bits of base metal, or paper – as well as bank money. These tokens become separated from their relation to any money commodity, or from the relation to the quantities of social-labour-time for which that money commodity itself acted as proxy. This illustrates the point that these tokens act as representatives of social-labour-time, not as representatives of gold or some other money commodity. If social productivity falls so that 1 million hours of value in commodities now becomes 1.1 million hours of value, the equivalent form of this value is 100,000 gold coins, each with a value of 11 hours labour, rather than 10 hours labour. Each coin is called £1, whether it originally comprised 10 hours of labour, or as now, 11 hours of labour. The price of each commodity, on average remains £1, and so there is no inflation. The same would be true if 100,000 £1 notes fulfilled this function. Indeed, assume that the value of gold does not rise, in which case, more gold would have to be put into circulation as currency, the value of gold would fall relative to other commodities, and gold prices of commodities would then rise. In effect, this is the same thing as if there had been a fall in the value of the money commodity, causing prices to rise. However, assume the currency takes the form of paper £1 notes. If the same 100,000 of them are put into circulation, they would continue to function as before. Each note bearing the name £1 would buy 11 hours of labour, in the form of commodities, including gold. There would be no inflation of commodity prices, and, in fact, the price of gold would fall. Each £1 note would now be a representative of 11 hours of social labour-time, as against the 10 hours of labour-time each previously represented.
But, the monetary authorities, seeing rising costs, do place additional currency into circulation. If the cost of constant capital increases, this is passed on into the higher value of output. The higher value of commodities increases the value of labour-power, reflected in higher wages. If a £1 note is fixed as representing 10 hours of labour-time, then a 10% increase in costs, and output value, must result in a 10% increase in the quantity of these notes put into circulation. Moreover, assume there is no increase in costs/values, but that wages rise, for example, because the demand for labour rises relative to supply. The consequence should be that, because the given amount of labour produces no additional value, the higher wages cause profits to fall. If the amount of currency in circulation remains constant, then prices could not rise, and so this fall in profits would ensue. However, because the state, and so central banks, are there to serve the interests of capital, which does not like to see its profits fall, when costs rise, as a result of rising wages, the central bank responds by issuing additional currency, which depreciates it, and so enables prices to rise. In fact, central banks do this also so that money wages rise, when otherwise they should fall due to rising productivity, because they know wages are sticky downwards. Firms increase their prices to cover the increased costs of wages, whilst maintaining their profit margins. The consequence is then that this further increases the value of labour-power, so now wages must rise again, which would squeeze profits unless prices rise, and so more currency is issued, and so on, in what becomes the familiar price-wage spiral seen in the 1970's.
Roberts theory of demand is also hopelessly wrong, as he says that demand comes from the new value created by labour during the year. That is the error made by Sismondi, Malthus and the Narodniks, and by Keynes, which leads to theories of under-consumption. Roberts version is essentially a rehash of Keynes modification of the erroneous Say's Law, which stems from what Marx called Adam Smith's "absurd dogma" that the value of output resolves entirely into revenues. Roberts simply repeats Smith's absurd dogma. The new value created during the year is equal to v + s, whereas the value of output during the year is equal to c + v + s. If demand, is only equal to v + s (i.e. revenues), its clear that it could not possibly be equal to c + v + s the value of output, and so there would necessarily be a crisis of under-consumption! But, as Marx makes clear demand is not, as Roberts claims, equal only to the new value created during the year (GDP) or v + s, but is equal to c + v + s, with the demand for c coming from capital not revenue.
At the end of his article, Roberts contradicts everything he had said at the start in relation to Monetarism, by saying that if the US Federal Reserve sees inflation rising too fast it will act to raise interest rates and reduce money supply!
His model predicts that US inflation could rise to over 3% by the end of the year, but its already at 4.2%!
No comments:
Post a Comment