Sunday, October 23, 2005

The Myth of Keynes and Sticky Wages

Most students studying Keynes will immediately assert with great confidence that the whole of the General Theory rests on Keynes' assumption of fixed nominal wages. This is the standard exposition in most macroeconomics texts and even pervades the thinking of most graduate schools. Asked why there is stubborn and intractably high unemployment in the economy, the standard answer is precisely the opposite of what Keynes wrote in the GT. The standard answer is that prices and wages are sticky downward. If wages and prices were more flexible, then there would be less of a problem of unemployment during recessions and the social cost would certainly be less. This is in fact the basis of New Keynesian thinking (as well as Neoclassical thinking) and it has become the mind set of our current generation of economists. All of this is now taken for granted despite the fact that Keynes was adamant in saying that flexibility of wages was not central to his assertion that high unemployment could exist over a period of years with no clear self-correcting mechanism to adjust the economy in the short run.

There are many economists who have staked their professional reputations on the notion of sticky wages. Thousands of pages of theoretical work have been written showing a myriad of reasons why that prices and wages do not move instantaneously to clear markets; despite the fact that it is in the best interests of both suppliers and demanders to have efficient markets.

Keynes was more savvy than this. The General Theory has at least two places where Keynes explicitly assumes flexible nominal wages. He uses flexibility of wages to show that a fall in wages may NOT lead to greater output and employment, if investment demand simultaneously drops. In the standard IS-LM and AS-AD model of modern macroeconomics what Keynes meant was that a fall in wages would no doubt lead to increased AS, but that the AD curve might fall if entrepeneurs felt that wages would continue to fall in the future. In such a case, firms would hestitate to invest now and decide instead to wait to invest when labor costs were at their rock bottom. An increasing AS and a decreasing AD might leave the economy no better off, with only a lower aggregate price and wage level.

In technical terms, the expected secular decline in wages and prices secularly lowers the expected nominal value of future profits. While the current price of capital has only partly fallen. This reduces the MEI and without the nominal interest rate falling equally as much, there is no incentive to invest. This is why that Keynes asserts that an expected secular decline in nominal wages will adversely affect investment. Thus, even with flexible wages, a fall in the nominal wage may lead to a concomitant fall in investment and therefore to no increase in output.

But, don't take it from me. Read the passages in the General Theory dealing with this subject and judge for yourself whether Keynes is justified in saying that the issue of flexible wages is beside the point. Section III of Chapter 11 deals with this subject and Section II part(4) of Chapter 19 talks about it at length. All it takes is a clear understanding of Keynes theory of the investment function, which has been a standard part of macroeconomics for over 60 years.

Remember this the next time someone tries to tell you that Keynesian economics just assumes fixed money wages.

Thursday, October 06, 2005

Keynes, The Monetary Tract, and The Inflation Tax

Long before the General Theory was formulated, Keynes wrote a short book called "A Tract on Monetary Reform". The book was published in 1923 and was devoted to discussing the issues surrounding inflation, which at that time had engulfed Germany and Russia. The book shows the remarkable knack Keynes had for illustrating complicated problems with simple examples. The examples he chose were always highly germaine to the problem he was discussing and were not constructed merely to show his intellectual prowess. There is no better example of this than his discussion of the inflation tax.

To begin with, Keynes was interested in showing that while it is true that inflation is a tax on an individual's real balances, one could show that this is equivalent to a tax on transactions or a "turnover tax". Modern authors of monetary theory seldom if ever show this equivalence, despite the fact that it is highly illuminating. Let's take Keynes own words now on the subject and see what they mean.

" Suppose that the rate of inflation is such that the value of the money falls by half each year, and suppose that the cash used by the public for retail purchases in shops is turned over 100 times a year (i.e. stays in one's pocket for half a week on the average); then this is only equivalent to a turnover tax of 1/2 per cent on each transaction. The public will gladly pay such a tax rather than suffer the trouble and inconvenience of barter with trams and tradesmen." (from Chapter 2 of the Monetary Tract).

As was stated above, few authors of works on monetary theory use this device to discuss the burden of the inflation tax. Indeed, the inflation tax is often simply written as equal to the product (the rate of inflation)(real money stock), with the inflation rate being taken as the tax rate on real balances and the incidence being dealt with via consumer surplus and the demand for real balances.

Keynes goes on,

" Even if the value of money falls by half every month, the public, by keeping their pocket-money so low that they turn it over once a day on the average, can still keep the tax down to the equivalent of less than 2 per cent on each transaction, or more precisely 4d. in the pound. Even such a terrific rate of depreciation as this is not sufficient, therefore, to counterbalance the advantages of using money rather than barter in the trifling business of daily life. This is the explanation why, even in Germany and in Russia, the Government's notes remained in current for many retail transactions. "

What remarkable insight, but we have come to expect this of Keynes.

The two quotes above seem to indicate that Keynes was using a formula to compute the inflation tax as a "turnover tax". But what formula is being used? Also, he uses a now anachronistic term for English pence and this becomes a stumbling block for some students of the text.

The second of these is easier to deal with. Keynes says in the second quote that the turnover tax rate had become less than 2% per transaction. He then makes this exact by saying 4d. in an English pound sterling. Now, at the time he was writing this there were 240 pennies (or pence) in a pound. That is not true nowadays. There are 100 pence in a pound now. But, at that time 240d. = 1 pound sterling. It follows that 4d./240d. is 1/60, which is less than 0.02. This is why he says that the tax rate is less than 2%. He means it is exactly equal to 5/3% or roughly 1.67%.

Now, how does he come by these numbers in the above two quotes? What is the formula he is using to compute these things?

The formula he is using can be written as:

(Turnover Tax Rate)(Number of Transactions) = (Rate of Depreciation of Money)

For example, in the first quote, we have the Rate of Depreciation of Money is 50% which is equivalent to a doubling of the price level with a stable stock of nominal balances. The number of transactions is assumed to be equal to 100. This implies that the turnover tax rate is equal to 1/2% and this is what Keynes gets himself.

In the second quote, we need to think carefully. Keynes assumes that the Rate of Depreciation per month is 50%. He then assumes that each dollar is held on average one day only, which means that the number of transactions per month is equal to 30. Hence, the turnover tax rate is 50/30% which is what we got before.

Monday, August 22, 2005

Read the General Theory Online

Some students, especially those in Taiwan, feel that the General Theory is a wonderful book, but perhaps not worth the cost of buying a bound copy. Naturally, I disagree with this, since a book can be taken anywhere and is always ready to yield its secrets, even without batteries.

Nevertheless, students may find the online version of Keynes' General Theory more to their liking. They can get this online resource at the following address:

One very positive aspect of this is that any particular passage of the GT can be printed out and worked at one's leisure. Another plus is that reading the General Theory online lets one instantly check definitions and references on the Internet. For example, Keynes often uses Latin expressions such as parri passu (see Chapter 10) and mutatis mutandis (see Chapter 20). Even when Keynes is not using Latin, his use of such terms as "prime cost" and "supplementary cost" (see Chapter 6) leave us uneasy with our understanding of what he is saying.

Additionally, you will find that this site above also includes the introductions Keynes wrote to the German, Japanese, and French editions of the General Theory. The French introduction is particularly revealing.

Of course, there is always the occasional teacher that will tell you that the General Theory has become a classic -- meaning that everyone knows about it, but no one reads it anymore. Which reminds me of something Donald Patinkin once said at a well-attended seminar at Texas A&M University. I was a graduate student at the time. Patinkin said that he was well aware that we had all read Marshall's Principles, so he didn't need to spend any time discussing background material. I turned somewhat red and realized that probably five people in the audience of one hundred economists and grad students had really read Marshall. Patinkin had just assumed we were as devoted to the classics as his colleages and students at Chicago. Later on, I realized the true importance of what he had said. Classics really do need to be read by people who typically go around dropping names like Keynes, Marshall, Samuelson, and Friedman in their economics lectures.

Sunday, August 21, 2005

What in the world is "user cost" ?

Most students who read the General Theory want to understand everything that Keynes wrote in his book. I can sympathize with this, since there is undoubtedly something depressing about reading so intensely but understanding only a part of such an important work in economics. Moreover, one can never be sure that some very important part has been overlooked or misunderstood---something which might clarify much of Keynes' thinking. Unfortunately the language of then and now has changed considerably. Keynes wrote the GT in 1935 and published it in 1936. Many of the terms, and even topics covered in the book, seem strange and archaic to us today. For example, almost no one uses the term "trade cycle" today, preferring instead to use the term "business cycle".

A particularly irritating part of the General Theory is Chapter 6 "The Definition of Income, Saving and Investment" and the Appendix to Chapter 6 entitled " Appendix on User Cost". I suspect many students get stymied with user cost at this point, which is a shame, since it contributes little to the main thrust of the book. Keynes spills a great deal of ink defining this notion of user cost and seems quite proud of his creation and its usefulness. This appendix nevertheless seems excessively complicated and nitpicking to our modern eyes.

So what is user cost and why is it important?

It will not be helpful to simply return to the General Theory and pore over the text one more time. After all, it is the text itself which is confusing. Fortunately, a very fine economist named Abba Lerner wrote a small paper which sets things straight for us. The small two-page piece entitled "User Cost and Prime User Cost" was published in the American Economic Review in 1943. After reading Lerner, the Appendix in the GT seems much easier to follow, at least to me.

So, what did Lerner say?

He begins by saying that prime cost is any cost that occurs because the firm is actually producing something. Prime costs are extra costs borne above the costs which occur when the flow of output is zero. These "extra" costs can be divided into three types: (1) extra factor cost, (2) extra costs from purchases from other firms--like raw materials, and (3)the extra wear and tear on machines (above what is replaced), which is very much like depreciation. Using Keynes' notation, Lerner compares these three where additional output is produced and no additional output is produced.

(1) F - F' = extra outlays on factors
(2) A1 - A1' = extra outlays on purchases from other firms
(3) G' - G = extra wear and tear costs on machinery

For example, F is factor cost of producing the output and F' is the factor cost when no output is produced. G is the value of capital if production takes place and G' is the value if nothing is produced. Note that some factors may be used to refurbish capital even when nothing is being produced.

Lerner then states that prime cost is the sum of (1), (2), and (3) above. That is,

Prime Cost = (F - F') + {(A1 - A1') + (G' - G)}
= Prime Factor Cost + Prime User Cost

He identifies Prime User Cost as the sum of the last two terms (A1 - A1') and (G' - G). By contrast, Keynes identifies his user cost with the following

User Cost = A1 + {G' - (F' + A1') - G}
=A1 + (G' - B') - G

where B' = F' + A1'. This last formulation is exactly as it appears in the GT. Thus, Keynes deducts from user cost factors being used to refurbish capital when such capital is not being employed to produce output. Lerner does not do this, but includes such factors in prime factor cost.

Now, it should be clear that Lerner's formulation is much simpler and intuitive. Keynes' user cost is more convoluted. However, the issue of how costs increase when output changes was terribly important to Keynes. It was this change in cost which was termed marginal cost (MC)and which was known to be equal to price(P) in competitive equilibrium.

Reading the Appendix to Chapter 6 reinforces the feeling we often get that Keynes thinking was not tied to some simple neoclassical model of the firm. Keynes wrote in the Appendix that even the great Pigou ignored these costs, thinking they were of secondary importance. Despite the fact that Keynes was right in emphasizing these costs and the effect they would have on his definition of income and the determination of price, we should not let this Appendix paralyze us in our reading of the GT. Instead we should recognize that the standard notions we have of cost are perhaps inferior to those of 60 years before. The old masters still have much to teach us that all the modern texts on microeconomics cannot give. It still pays to read the masters whenever we can.

Saturday, August 13, 2005

Mercantilism in the General Theory

The history of economic thought during the late 18th century is a story of the unqualified repudiation of mercantilism. Mercantilism was a system of thought in economics that identified increases in national wealth with a positive balance of trade. Policies which led to balance of trade surpluses were to be vigorously pursued. Such policies would lead to a steady accumulation of gold specie which was identified with national wealth, output growth, and prosperity.

Today, the same faulty mercantilist sentiments are often expressed by individuals and the media when they speak in glowing terms about the large amounts of foreign exchange of a particular country. For example, one often hears people (who should know better)in Taiwan speak with great pride and sense of accomplishment that Taiwan has the third largest amount of foreign exchange reserves in the world. They express this view despite the fact that Taiwan is not anywhere near the third wealthiest country in the world. They express this view despite the fact that Taiwan's economic growth has recently fallen to historical lows. These mercantilist notions play well to audiences determined to see greater protectionism and nationalism. Stocks of gold have come to be replaced by foreign exchange reserves and special drawing rights at the IMF. Nevertheless, it remains mercantilism all the same.

Despite the manifest failure of mercantilism to withstand the criticisms of 18th and 19th century Classical economists, Keynes found space in his General Theory to defend in part mercantilist policy prescriptions. His purpose in doing this was to show that early mercantilist writers had an intuitive feel for the value of maintaining low interest rates, finding an outlet for the over-production of goods, and increasing the money supply(gold). These three goals conform exactly with Keynesian theory as expounded in the General Theory.

Keynes presents these highly controversial views in Chapter 23 of the General Theory. Oddly enough, they are rarely discussed in macroeconomics classes nowadays. Lacking central banks, these early economies required stable policies which would ensure increasing money supplies to accompany economic growth. Trade surpluses and concomitant gold inflows were essentially the only way to do this.

Once again we have an example of Keynes' unconventional thinking backed by an enormous, very persuasive, and complete theoretical superstructure. No wonder Keynes' General Theory represented a revolutionary change in economics and set in motion an entirely new direction of research. No wonder that we stand in awe of someone who could martial so many different strands of evidence to support his views. Rarely do we see this kind of rhetoric today. It is indeed a tragedy that many economists today have completely lost their ability to argue their points forcefully. They content themselves with the delusion that talking to each other in ceremonial scientific trappings can make up for the obvious erosion in their ever-dwindling audience.

Friday, August 12, 2005

The Danger of a Sudden Rise in Precautionary Money Demand

With the advent of his General Theory, Keynes is thought to have focused attention on the importance of the speculative demand for money and its role in determining interest rates. This part of the demand for money has subsequently received the greatest degree of notoriety. John Hicks claimed that it is the small speculative or voluntary part of the demand for money that is really important to monetary theory (See his Two Triads:Lecture I, last paragraph in his Critical Essays in Monetary Theory). The precautionary motive has been typically ignored or downplayed by most economists,although Hicks did lump it in with the speculative part.

However, Keynes makes reference to how that a rise in uncertainty, not only about the normal level of bond yields, but also about the economy in general, might cause a sudden and disastrous rise in this precautionary demand for money. It was this rush for precautionary money (in fact, base money) that caused the meltdown of the banking system in 1933 in the US. It was to happen again, over and over, in countries experiencing capital flight. Massive sales of non-monetary assets in preference for cash, intended in turn for quick conversion to a more stable currency. In a sense, this process is like people facing a fire in a theater, where the door to the lobby must be passed before exiting the door to the theater itself. The rise in precautionary money demand is extremely dangerous since it can lead to a contraction in the supply of money just as the demand for money is expanding. The ever-widening gulf between supply and demand in the money market can create a bank panic and a severe contraction. It is the fact that both demand and supply of money are moving in opposite directions that makes it problematic for the central bank to halt the contraction by simply providing reserves to the banking system. There is a psychological aspect to the problem which is not usually fully appreciated. (e.g. Albert Hettinger, Jr.'s thoughtful and gingerly written comments at the end of Friedman and Schwartz's The Great Contraction politely criticizes their analysis for just this reason--not enough emphasis on psychological factors).

Keynes never failed to appreciate the role of psychology, tempered by economic and institutional constraints. The General Theory is full of such examples with respect to the demand for money. For example, in Chapter 22 Notes on the Trade Cycle, Keynes clearly states -------"Moreover, the dismay and uncertainty as to the future which accompanies a collapse in the marginal efficiency of capital naturally precipitates a sharp increase in liquidity preference" -------and it is clear that he is speaking of the precautionary component, as opposed to the speculative component of money demand. Even here Keynes is so focused upon his new toy -- liquidity preference -- that he apparently forgets that this rise simultaneously raises the currency/deposit ratio and the reserve/deposit ratio, thus contracting the money supply.

The above shows that it is of paramount importance for central banks to take pre-emptive measures to head off any chance of a sudden and catastrophic rise in the precautionary demand for money. Certainly, this responsibility far outweighs any admonition to maintain a low and steady rate of inflation.

Thursday, August 11, 2005

What was so "general" about Keynes' General Theory ?

I once had a teacher in college who told me that he had taken a final exam in his own university days where his professor had asked students the following question: What was so "general" about Keynes' General Theory? That's a great question. Now let me try to answer it.

Keynes argued that Classical macroeconomic theory was full of special assumptions--assumptions which he felt were not at all realistic. Classical thought, according to Keynes, assumed that adjustment in three variables would generate economic recovery from recessions. Classical economists asserted that during times of inventory recession wages, interest rates, and prices would all fall and return the economy to high production and full employment. Keynes said this might be true sometimes, but a more general theory would have to consider cases where this would not occur. Let's look at each variable separately.

First, it is certainly true that wages will fall with high unemployment, but this fall might well be accompanied by expectations of a further fall in wages. If firms expect a further fall in wages, they might very well hold off on their investment plans. The reason is simple. Falling wages and prices in the future meant that the medium to long term stream of nominal profits from investment would also be lower. This would lower the nominal MEC and hence reduce investment. Only if the nominal interest rate fell sufficiently would there be any reason for investment to increase. In general, falling wages accompanied by an expectation of further falls in wages would dampen investment and lead to a reduction in aggregate demand. Falling wages would not lead to increased employment and general. Only in the special case where the fall in wages was expected to halt or reverse its course, or the nominal interest rate fell sufficiently, would there be a rise in investment and an increase in production and employment. (See Chapter 11 and 19 of the GT)

Second, a reduction in the interest rate, due to excessive household saving and an "over-production" of consumer goods, was thought by Classical theorists to raise investment demand and revive the economy by increasing the production of capital goods, machines, structures, and the like. Keynes asserted that this was only true if saving and investment alone determined the interest rate. He pointed out that this was a special and unrealistic assumption. It was more likely that output would adjust downwards, rather than the interest rate, if there were excessive saving. Furthermore, investment demand would most likely shrink even as the interest rate fell, with the result that the quantity of investment might also fall. (See Chapter 14 and Keynes' only use of a graph in the GT)

Finally, falling prices during an inventory recession might not increase spending and return the economy to a position of full employment. Certainly falling prices increase the real stock of money and thus place pressure on the interest rate to fall. However, Classical economists made the special assumption that the rate of interest would always fall sufficiently to restore full employment. Keynes declared that if there was a liquidity trap or if investment demand collapsed due to falling prices and a loss of confidence, it might be impossible for the interest rate to fall below a certain level. Arthur Cecil Pigou later countered Keynes theory by declaring that falling prices would generate a real balance effect which would stimulate consumption and hence aggregate demand. (See Chapter 17 of GT for Keynes on the liquidity trap)

According to Keynes, Classical theory was not illogical. For him, the trouble with Classical thought was that it relied too heavily on very special and unrealistic assumptions, thus rendering it largely irrelevant.