Wayne McMillan suggested me to read Mitchell and Wray's new book ''Modern Monetary Theory and Practice'' which is in the final process of redaction. The following is my rough impression of their new book in special reference to the notion of effective demand.

In Chapter 11, the authors treat Keynes's theory of effective demand. Adoption of the concept Employment-Output function is for me doubtful. Principle of effective demand tells in my interpretation that it is the effective demand which determines the amount of production and therefore the amount of employed labor. In this sense, the real function, if it reflects any causal determination between variables, is the function which determines amount of labor by the effective demand. This function is named by Keynes Employment Function: N = F(D). It will be better that Mitchell and Wray will focus on this aspect of the "aggregate supply function" instead of using misleading Y = αN formulation. Mitchell and Wray are still thinking in the framework of symmetric equilibrium between demand and supply.

What I want to suggest is to reorganize all discourse into the process analysis framework. In this analysis, the direction of determination, or which variables determine which variables, are important. In the case of effective demand, effective demand is determined by the spending of households and firms either for consumption or investment among the income they get previously. Then the effective demand determines the amount of production and by consequence amount of labor employed. The employment gives rise to wages, or households' income. The profits of firms will be determined in the same way. Then, the households and firms will engage in their consumption or investment activities. This occurs in a circular process and this is not an equilibrium state.

I emphasize the necessity to formulate effective demand in the process analysis because I believe this is the unique way to incorporate the real function of money. In the capitalist monetary economy, the money buys goods but goods do not buy money. It is the demand, equipped with money, that will purchase the commodity (goods and services) and the firms respond to this demand. Symmetric equilibrium framework ignores this basic facts and loses to grasp the real process of monetary production economy. In this sense, I think Mitchell and Wray's new book goes half way to the construction of coherent heterodox alternative. I also think it will fail to show the readers the real essence of the function of money in modern market economy.

Interesting question, Yoshinori. However the Keynesian equations based on supply-demand theory are only follow the usual assumed forms (and thus have an intersection and are analytically tractable) if they are a negligible fraction of an economy. If you aggregate an entire economy and attempt to write such equations I think you will find there is no guarantee of intersection at all. If there is no intersection, the economy keeps increasing or decreasing output looking for an intersection, with no guarantee of an endpoint until at extremes a severe non-linearity is met in the functions.

To be more specific, in the aggregate consumption demand is some approximately proportional function to the number of employed workers N. When looking at a fully aggregated economy, we can remove the confusion of pricing and just realize that the entire production is consumed up to some maximum amount. (If you use a pricing model, you simply find demand pricing drops precipitously at this point). So what you find is both the supply and demand rise with the number of employed workers N.

To examine the conditions for there being an intersection at all, then we look at the average production per worker. Is this more or less than the average demand per worker? If Production is less than average demand, then there is absolute demand and society will invest capital to make production more efficient to meet demand. This characterizes most of human history, and in particular the era in which Keynes formulated his theory. His theory led to effective policy to make cheap capital available and build productions, so it was an unqualified success at the time.

However, if a society reaches a point due to productivity growth (presumably due to past investment) where aggregate demand is less than average production, and per worker demand is less than per worker production, then employment will fall precipitously and there will be no intersection anywhere of the two curves. One worker makes more than she needs, and no one else is employed to buy the excess. If one tries to "stimulate" such an economy, productivity grows further and in the limit a machine produces all output and no one is employed. We need a new theory that deals with this situation. The only method so far proposed is to keep thinking up new products, but that is running out of steam, not to mention that it is not really analyzable in theory.

Robert, thanks for your stimulating post. I read the introductory part of your book and noticed that, although our ages are quite different, we have passed similar high school days, for I am also interested in Four Color Problem in those days. I am trained as mathematician and changed to be an economist at the age of about 30 but still are fascinated in Mathematics and Quantum Mechanics and others.

I will develop what I think to be a good theory of effective demand later. For the moment, let me remind that I am not satisfied with the Keynes's formulation of effective demand. I am against the standard model with aggregate demand function and aggregate supply function.

My objection is conceptual. I don’t suppose that they intersect at any point. I doubt the very existence (or the concept) of two aggregate functions with total number of employed workers as independent variable. There are no such functions. There are no causal chains which connect number of employed workers to a state of demand and a state of supply. This does not imply that the number of employment is related to demand. On the contrary, if there is a sufficient demand, the firm managers will employ new workers. It is the effective demand which determines the number of employed workers. This relationship is expressed by employment function. Keynes explains that the employment function is only an inverse function of the aggregate demand function, but this explanation should be reversed. What Keynes saw as aggregate demand function is but an inverse function of the employment function.

This crucial difference is put off and robbed out in the standard equilibrium framework.

You are heading in an interesting direction, Yoshinori. You realize I'm sure that by changing so many assumptions it will be harder to gain traction with others. If you would like a complimentary copy of the book, send me a private message with your address.

I had a new idea this morning which I'll run by you. I want to explain why sometimes productivity results in expansion and hiring, and sometimes not. My suspicion is that it relates to pricing and thus to money basis. In theory of course world demand could be saturated but there are not really too many outputs for which world demand is saturated, possibly none outside developed nations.

So I propose looking at incremental revenue. This somewhat violates my previous comment about necessity of looking at the whole economy, but sometimes it is necessary if one doesn't understand the whole to just analyze a delta to the current situation.

Now I consider two money bases. In reality we have a complex regulated basis which is somewhere in between, but for theoretical discussion I compare a commodity basis, which in theory free trade floating currencies are, and a fixed basis in which money is a contract for delivery of a slowly growing or static commodity like gold. (Aside: for Southern European countries, using the Euro as a basis is for them a lot like using gold).

For the commodity basis money, money is a contract to deliver a quantity of generic production. If a business makes a productivity improvement, then the added production brings in additional revenue in absolute terms, and an increased rate of profit due to the productivity. No rational businessman would turn this down by laying off workers.

for the fixed basis money, in the long haul no additional revenue is brought in by the increase in productivity. The total supply of money is relatively fixed, and other businesses will be increasing productivity as well - they all have the same environment and motives and technology - so prices decline approximately in proportion to the productivity increase all around and no new revenue results.

That implies that in the second scenario, the only profit motive for the rational businessman is to lay off workers and use the productivity to produce more or less the same volume of goods as before.

The thing that puzzles me is that it seems in this "thought experiment" that a substantive difference in the economy results from an artificial difference in the choice of money basis, and I wonder if I am overlooking something. What do you think?

It is good to think of business person. But it depends much on which of ranks you are thinking of. If we want to talk about rudimentary supply attitude of firms, the good person will be a factory manager who is in charge of deciding how much of a product and of which products the workers should make today or this week. What is, do you think, the most important factor in this decision making? It must be the estimated sales volume for today or for this week. But, these estimations are most of the time a question of guessing and the managers have to appeal to a kind of trend prolongation using the past sales volumes.

If sales go well, the manager will increase the production and in case of manpower in short he or she will try to recruit new workers. If we sum up these decisions for all firms, we will get the total number of employed workers.

If you know the total amount of demand the people want to spend today or this week, and if you know how they are divided among various products of various firms, then you can get a function in which total effective demand determines the total volume of employment. This is what Keynes names employment function (See Cahpter 20 of his General Theory). But there is no other way round. This is the reason why I propose to abandon the concept of aggregate supply function.

Productivity and monetary situation may change the form of the employment function but you cannot skip observing this elementary process and jump to an analysis of aggregate variables. In that abstract way, I don’t think we can arrive at a good conception of how the principle of effective demand works.

Hi Yoshinori, I believe I understand the sound reasons why you make this effectively "instaneous" or near term demand prediction argument (based on trends or whatever), and I think you correctly identified that some type of mid-level manager, or factory manager, will be making this sort of decision.

However, this defensible, local view does not fit recent data. Demand increases but employment does not.

With that in mind, my specialty is looking for, and usually finding, longer range "circuits" feedback or unexpected effects. I not only intuit this from my own logic, but I read of top level planning and strategy far above the factory manager. I also experienced this in a large organization of which I was a part for 40 years. Orders would come from the top which did not support the logic of the factory floor manager. These would often be generic "policy" orders to close facilities and become more productive and reduce hiring even in the face of demand. So I know that this does happen from experience.

While I'm not privy to the top level conferences, I can read the same books they do, and the same financial news coverage, and I believe the reasoning at the top level is as I have described it, and the reason these guys (CEOs and the like) draw down the big bucks is because they do have surprising effects on organizations.

Now if you want just to make a mathematical theory that will appeal to fellow economists, by all means proceed as you are going. If you want to describe the real world, and the lack of hiring in the face of demand, while preserving some semblance of economic reasoning, then you must find why there is [apparently] no hiring incentive in the face of demand at the current time, which means it must be in the current context or environment.

So that leaves many possibilities and you may find others than I named. But my suggestion is to look at investment in productivity because that is one of the most prominent trends in the 21st century, and to the extent behavior is empirically divergent from existing models, should lead to new theoretical models.

I would include under the broad term productivity both the offshoring and outsourcing and hiring of temps, which is a kind of productivity trick which increases product per dollar but not product per labor hour (in might even decline in that regard), and also productivity for automation. Both require long term planning and investment, beyond hiring, which are not the purview of the factory manager, but of the CEO. The CEOs have plainly issued orders that demand is to be met by automation and temps and outsourcing.

Another possibility is that CEOs have decided the demand is transient and they will ignore it. This has been widely speculated in the media and there may be something to it. But I did not address it in my previous note. It is pretty subjective. I only addressed the more easily quantified currency incentives. And really I was just asking your opinion on that, more than making a specific proposition for a model. I am wondering if indeed a money basis can have an effect on the real or perceived reward from demand? I'm not sure I have thought of every angle.

Robert, as you have indicated, productivity is rather a vague concept. Suppose you are observing a production of a product A in a factory, the input coefficients can be measured. Let them be a0, a1, a2, ..., aN per unit of product A. The coefficient a0 is the labor input and others are input for each input good. If w is the wage rate, and p1, p2, ... , pN are prices, then the unit cost of product A is

u(A) = w a0 + p1 a1 + p2 a2 + ... + pN aN.

If unit cost u(A) diminishes during a certain lapse of time, we can say that productivity increased. But this kind of measurement depends on the wage and prices.

Complication occurs when some of coefficients decrease whereas some others increase. Without appealing to wage and price system, we cannot even tell if this change of coefficients means a gain in productivity or not.

We can concentrate our attention to a specific input. For example, we may talk about labor productivity, which is the inverse of a0. The labor productivity increases sometime tremendously. My colleague Fujimoto observed that in many Japanese factories, labor productivity increases 15 to 25 percents per year and this pace continues for 3 to 5 years without any major investment in installations and machines. In an IC factory, the labor productivity increased 10 times within 4 to 5 years.

If there is this amount of labor productivity increase, there is no necessity for the labor power to increase when the annual sales volume increases 10 percent or so. Indeed, we have an equation

l(t)= y(t) /a0(t),

where l(t), y(t), and a0(t) are the labor time, production volume, and the labor input coefficient at time t. The necessary labor power increases only when

Δy(t)/y(t) > ΔPr(t)/Pr(t)

if we write Pr(t) = 1/a0(t), i.e. the labor productivity..

You talk about productivity gain by capital investment. But what kind of productivity are you thinking of? Is it a gain of total factor productivity? Are you thinking of labor productivity?

If you are only observing macro data, it is doubtful if you can arrive at finding a good stable relation among macro variables, like total labor input and capital input. Anything can happen to these variables. You should remind the Phillips curve debate in 1970's. Until the end of 1960's, Philips curve seemed fairly stable but after 1970, it became fragile and we could no longer deem it a determined function.

This is one of reasons why I think that a really new theory of effective demand is necessary now.

Hi Yoshinori, yes I agree productivity is an ambiguous concept, and since I graduated from college and had to find a job in the 70s, I have first hand experience with the lack of long term validity of the Phillips curve. : )

I was thinking of capital investment in automation mostly. Traditionally most kinds of capital investment, like even old style factories, were for the purpose of improving productivity over trade-crafted goods.

I certainly agree an improved theory of effective demand is needed.

Mains Points for the Reformulation of Keynes's Theory of Effective Demand (1)

There are several points which should be made clear, for the theory of effective demand requires reformulation of economic theory from the very foundations.

First of all, we have to liberate ourselves from the neoclassical price theory. It started as marginal theory of utility and was systematized by Alfred Marshall. A modern form of the neoclassical price theory is Arrow-Debreu theory on the existence of competitive equilibrium. Debreu wrote a concise book "Theory of Value." It is a wonderful mathematics book. I do not deny it. The trouble with these neoclassical theories is that they assume an equality of demand and supply for all goods and services. The neoclassical researchers call this state equilibrium. According to these theories, economy can be analyzed as it is always in equilibrium. An equilibrium can have an excess supply but at that time the price is 0 (rule of free goods).

If we admit these theories, it is impossible to construct effective demand theory, for the lack of effective demand causes involuntary unemployment. In the neoclassical theory this kind of involuntary unemployment cannot exist by the very construction of the theory. If there is unemployment, it means that those who demand work but not employed are asking too high a wage (rule of free goods [in this case service]).

This was the reason why a research program called "micro foundation of macroeconomics" failed. This research program was proposed by Clower and Leijonhufvud (in 1960's), and followed by Malinvaud, Benassy, Negishi, and other Keynesians. But this research program became soon (in 1970's) anti-Keynesian and offered a good reason to deny the Keynesian theory.

As a theory is only dislodged by a theory, the first question we have to solve is to find or construct a new value theory on which we can base Keynes's theory of effective demand.

Hi Yoshinori, I propose a test case for your efforts to construct a theory of effective demand. Maybe it is not exactly the case you had in mind, but it is an important case in the world illustrating a conundrum between demand and unemployment. It is a case where there is both demand and unemployment. I do not have the facts to decide if it is a resource issue or a currency issue, but they should be easily obtainable. The question arises from a radio interview with two women in Bangladesh who left village life to work in (sometimes dangerous) garment factories.

They left the village because in the traditional way of life, for a long time, I presume we could assume equilibrium, people were borderline starving. Now they send a small amount of money home which provides ample chicken for the pot, and the younger one has figured out she does not have to marry a man chosen by her family for apparent income (often deceptive), and does not have to marry at all.

The conundrum is this. The transition apparently wasn't possible without trade. But trade should have nothing to do with it. The people of Bangladesh should be able to at least feed themselves, or no one has any business living there except to extract and export natural resources. Simply a labor colony is not sufficient rationale.

Why weren't the villagers of Bangladesh able to obtain chicken? (or food generally) Surplus labor was available in more than sufficient quantity to raise chickens. (or food generally) The labor was not able to purchase chickens. Viewed without the lens of currency, there was plenty of demand (labor which wished chickens). Economic incentives internal to Bangladesh failed to organize supply and demand, and failed to organize labor to meet demand.

Now you see the reverse problem in Japan. Economic incentives fail to balance labor supply with demand. In both cases there is excess labor. In the former, it goes unused. In the latter, labor or organized for production, but the distribution (demand) fails.

A principle difference in the two situations is the level at which demand is satisfied. I am thinking of the Maslow needs hierarchy - 1. food & shelter, 2. security & safety, 3. love & belonging, 4. esteem, 5. self-actualization.

In Bangladesh, the need level was 1. If the country is viable at all, this need should have been met. It is possible level 2 if lacking can suppress 1 however.

Japan has for much of history been obsessed with level 4. In that case it is employment itself that meets the need (for esteem), and so demand for the goods and services produced is not even in the equation. It might be that any conventional economically formulated theory fails in such a situation.

Robert,

thank you for your unexpected comment. To think of a "text case" maybe very common in physics but, in economics this kind of thinking is a rare habit. When we talk about "testing" in economics, it normally means test a proposition or alternatives on econometric base. In most of the cases of theory making econometric testing is quite difficult by the lack of suitable data. What you have proposed is a quite different from those tests. It is a proposal to make a thought experiment (or Gedankenexperiment). This important way of thinking was lost or was not become common practice in economics. You have reminded me a very important teaching of modern physics.

I will try to think about your "test case." But as I have an urgent obligation to fulfill this week, the response will be in the next week.

Glad you found it useful, Yoshinori. As is obvious from my profile, my other passion happens to be physics. :D I am about to upload a paper I've been trying to write for many years giving a derivation of the Lorentz transform and by implication all of Special Relativity, without using the relativistic postulate. The key to finally being able to do the work was the absolute clarity that the Lorentz transform is a geometric mean of (1+v/c) and (1-v/c) and though I had stared at the formula for 40 years, I did not see that until I spent a few years working on problems in Finance theory. So each field has something to offer the other.

Robert,

your example requires many points of clarifications and explanations. I will present what I understand on the fundamental difference between Japan and Bangladesh later, but first of all, I have to confess that your last comment was out of my imagination. But, it must be the very important point when we want to analyze employment / unemployment problem. Yes, employment itself must be something which meets the need (for esteem) in Japan, whereas food and shelter must be more important demand in Bangladesh. Normally, in economics, employment means toil. Wage is the compensation for this pain and fatigues. Then, in the neoclassical formulation (or the Second Postulate of the "classical economics" after Keynes) the demand for employment decreases when the real wage rate decreases. When employment means esteem (necessary to keep self esteem, for example), demand for employment becomes quite independent of the level of real wage rate. This must change the discussion very different.

Keynes himself denied the Second Postulate. "This postulate is compatible with what may be called 'frictional' unemployment." "(T)he postulate is also compatible with 'voluntary' unemployment" (General Theory, Chapt. 2). But it is incompatible with what Keynes defined as involuntary unemployment. This is the point the New Classical Macroeconomics has forgotten. For example, real business cycle (RBC) people do not admit involuntary unemployment. When unemployment increases in depression, RBC people interpret that workers prefer leisure than to work. So the question of the Second Postulate is quite controversial even today. This is one of reasons why I want to reconstruct theory of effective demand. But, this is not the point I want to discuss now. Let us return to our test case, i.e. the Bangladesh and Japan case.

You made an interesting comment on the Bangladesh village case. Your wording was "viewed without the lens of currency." Bangladesh village case is more easily understood as the question of subsistence economy without money. Early development theorists (for example, Arthur Lewis) thought that the 20th century underdeveloped economy as mixture (or in a more appropriate terminology dual economy) of industrial capitalist economy and subsistence economy. In the subsistence economy, people lives mainly in rural areas depending on traditional cultivations and raisings. Because the peasants do not have enough capital, they cannot engage in industrial works without employed by industrialists, who have capital money and knowhow’s. Then, in order to enhance family the woman has to leave her village and send a part of her wage to the family. If she has a small amount of money (a few thousand dollars), she can buy several chicks and food for them and she may feed her family more easily in her village. This was just the situation where Muhammad Yunus started his Grameen Bank, or the first enterprise of the Micro finance system. With the small amount of capital money, peasant women could change their lives.

We cannot say that it was the lack of effective demand which caused unemployment in Bangladesh. This must be the case of capital accumulation problem for Classical Economists. (I mean in this case Smith and Ricardo and other classical economists before the arrival of neoclassical economics. Keynes called this neoclassical economics "classical economics" and spawned many confusions. In my opinion, it is important to distinguish Classical economics and Neoclassical economics.) Normally effective demand is interpreted as demand backed with money. Bangladesh village woman lacks her money and it was the very reason that she had to go to a city to find her employment.

Japanese case is different. Of course, there are many poor people who are deprived of enough money income. But, at the same time, there are many people who have non negligible amount of finance assets but do not want to dispense their money. Reasons for each person may be different. Some think of their retired days and want to conserve their wealth. Others are satisfied with their present level of consumption and have not an eager interest in spending their money.

If some fundamentally new products such as PC's in 1990's appear, they may be happy to buy those products. But, for the moment, they are rather satisfied. If people earn money and if they do not spend them, money economy does not work. What consumers save cannot find enough opportunity for investment (in real economy). The rest of the saved money must stay in financial or asset economy. As the Japanese finance enterprises cannot find a good opportunity to investment in Japan, the money may go to New York and will be dispersed all over the world. If we can find good opportunities somewhere, this may work for some time. But many of these opportunities are not very safe and stable and may produce a crisis like Lehman shock.

Thus Japan's problem is virtually common to all advanced countries. There may be some differences between countries. For example, U.S. enjoyed a high rate of consumption growth, partly due to excess consumer loans and excess housing loans.

The problem of lack of enough effective demand is not only connected to unemployment but also to the growing "financialization" of the world economy.

Mains Points for the Reformulation of Keynes's Theory of Effective Demand (2)

My second point for the reformulation of the theory of effective demand may be most controversial. I propose to distinguish two economic domains: real economy and asset economy (or financial economy). The two are closely connected and it is sometimes difficult to distinguish them. Even though I believe it is very important to distinguish these two domains.

This kind of distinction is quite popular in newspapers but not very common in pure economics. Keynes himself made a similar distinction in the mode of monetary circulations in his Treatise on Money, i.e. industry circulation and financial circulation. Although Keynes abandoned this distinction in the General Theory, these two domains of monetary circulation are useful point of view when we analyze total economy.

Real economy is the domain of production, investment (to increase means of productions and productive capacity), and consumption. We may say production economy. If we say real economy, this does not mean we think of barter economy without money. Real economy is a part of monetary economy and is moved by money. Most simple formulation in this domain is the next saying: Money buys goods (and services) but goods (and services) do not buy money. We have to establish a "monetray theory of production", as Keynes tried to do in the first draft of his General Theory.

Asset economy or financial economy is the domain of economy where financial assets and some physical assets are exchanged. As the finacialization goes on, weight of asset economy is increasing with proportion to real economy. I do not have an intension to denounce asset or financial economy. What I want to clarify is the structure of world economy which compels the economy to propel the financialization.

The need to distinguish two domains lies in the fact that saved money is not necessarily invested in the real economy. Major part of savings may circulate in the asset economy for a substantial length of time. Then the famous formula that requires S = I should be revised in view of this reality.

Immediate consequences which we obtain from the distinction of two economies:

Before going to the third point, it would be convenient to give some of its immediate consequences of the second point of the reformulation. If the total economy is divided into two different domains of economy, it is reasonable to differentiate rules and properties which govern each domain. For example, a law which holds in one economy may not be true for another economy.

In a real economy, we can assume that prices are determined by markups, i.e. adding some conventional margin to the unit cost. In an asset economy, the price determination must be very different from this. The price of a kind of assets depends totally on the current supply and demand of the specific asset. This is rather similar to Walrassian organized market but contains a crucial difference. In the Walrassian market, the prices are set when the demand and supply of all kinds of commodities become equal. In the actual finance or asset market, prices are determined if demand and supply comes to balance for a specific kind. Another important difference with respect to Walrassian formulation is that equilibrium does not imply the subsistence of that price. The prices of asset economy constantly fluctuate and do not come to a fixed price. Even the volatility changes from time to time. In some cases, volatility is an important factor to guess the level of prices.

Effects of exchange are also remarkable. No changes occur for the existent quantities of assets. When the prices changes, there may be some distributive effects for those possessors of assets. Those who have an asset lose when the price of that asset decreases and gain when the price goes up. The total amount of an asset remains constant through these operations. The amount of an asset changes only when stock or bonds were issued. Or, in the case of real estate, it changes when they are created or constructed.

What happens in real economy is quite different. The production gives birth to a new good. Consumption reduces the amount of goods you have bought. You can increase your stock when you buy a new good. Normally the purchase takes a form of an exchange. But in this case, when you purchase, it increases the volume of production and working time and working produces new values.

The logic of an exchange changes differs so much according as it occurs in an asset economy or it occurs in a real economy. It is reasonable to distinguish two economies when we want to analyze what happens in an economy. This is of course one of the main reasons why we should distinguish two economies.

Hi Yoshinori,

I have been traveling for Christmas, and for several days without email. I see you have made many posts and some quite important points.

I see that you understand all of my rambling post about Japan and Bangladesh. Later you state: "Japan's problem is virtually common to all advanced countries." With this I agree, without necessarily specifying precisely all the nuances of what "Japan's problem is." However, this is almost universally denied in the United States, even as the U.S. enters the early phase of something like the long Japanese decline. It is only less sudden here because there is not an equivalent trading partner outside the U.S. to impose the sudden currency appreciation on the U.S. as happened to Japan in the mid 1980s. Current pressures in this direction come from two directions, Japan which in attempting to devalue its own currency pressures the U.S. currency upward, but the Japanese actions are too little too late as they have been for 25 years .... and secondly Europe which is busily sabotaging its own currency, but not deliberately, only through political and economic incompetence, and so it is not especially rapid. China also figures into the picture of keeping U.S. currency valuations high, and so does the U.S. electorate, which is not nearly as savvy as it was during the 1980s when it allowed the devaluation of the U.S. dollar. The problems are not really currency problems, though, they are structural problems. You would say they are problems of demand, and as you know already I say they are problems of productivity. But I acknowledge some demand problems and that leads to ...

You state: "If people earn money and if they do not spend them, money economy does not work." This is a feature particularly acute when wealth accumulates in a smaller and smaller fraction of the population. The very wealthy have a lower overhead, and therefore a larger surplus for which they have the option of hoarding. By hoarding I do not mean putting money in the mattress. For the wealthy, re-investment is hoarding. They expect returns from re-investment. The re-investment increases production, but the problem is the wealthy are not consuming with the money they have, so in the situation in which there is not a production deficit, their re-investment is counter-productive (odd usage of the word in a meta sense).

World economic problems fluctuate between production and distribution. I am currently reading Tenney Frank's Economic History of Rome. It is uncanny the similarity of the evolution of the Roman economy to today's world. Especially the U.S., which consciously emulates Roman Republic in political structure. They went from a surplus of grain production (which U.S. currently has) to an excess of military power, directly as a result of grain productivity (again parallel to U.S., farmboys into soldiers). With over cultivation of thin volcanic soil, grain production fell, driving conquest to supply grain. This is an oversimplification, and largely Frank's conclusion upon which I rely. But it seems to ring true. They gradually and fitfully migrated from a society of independent farmers to a plantation society which was able to more efficiently utilize the degraded land, now fit more for pasture and olive groves. The unemployed collected in Rome and were supported with the corn doles. So Rome had this unemployment problem for hundreds of years. The decline in Japan was characterized by the loss of "lifetime employment." The current situation in the U.S. is characterized by the "permanently unemployed." Whether to call this voluntary or not is not a straightforward question. If you had lived at the U.S. standard of living for generations, and your employer offered you your same job in India at the prevailing rate there, and you declined, would this be voluntary unemployment? You can't find work in the U.S. Or anywhere close to the standard of living upon which you based your economic investment in a STEM degree. The employer that did just this a few years ago was IBM, and the employees targeted were engineers.

Just some things to think about. I have to go visit with relatives now. On Bangladesh, you have described accurately the current situation and referred to prior theory, but I do not see the explanation of why the village life was breaking down and they could not feed themselves. It seems similar to the breakdown of the Latin farms, or is it? Is it intrinsic, exhausting fertility of the land? Overpopulation pressure? Or industrial encroachment pressure?

Welcome back, Robert. You must have you enjoyed your travel without e-mail. Your last post contains so many points that I cannot answer or refer to all of them. I want only to mention what seems most important: the concept of hoarding.

You have written: By hoarding I do not mean putting money in the mattress. For the wealthy, re-investment is hoarding.

I totally agree with you. This is the most important and yet most neglected point. The hoarded money is invested in a kind of assets. When it is invested to a newly issued stocks or bonds, and when this amount of money is invested to expand production, to the improvement of production installations, we can say it is the classical type of investment (investment in real economy). But the point is that this classical type of investment occupies now a small percent of all investment. Major part of investment goes to circulate in the asset economy or finance economy. When the money circulates in the asset economy, it produces nothing. Money changes the hands but it does not increase assets. Money circulation may stimulate price changes and sometimes may induce a bubble.

All the money held in asset economy for a certain period of time is to be named “hoarding”. Keynes did not notice this. He supposed that any money invested in finance market will be finally invested in real economy. But the money can stay quasi-permanently in the asset economy. If this amount of money increases in volumes, it will not stay gently and induces price volatility to increase. Story changes much from that described by Keynes.

Wish you a Happy New Year.

Hi Yoshinori,

Yes I did enjoy some time without email. I accomplished completely impossible things . . . like patching a blown out radiator hose and then driving 900 miles. :D))

I am glad to see you distinguish between investing in new production assets, and old production assets merely changing hands at different prices. I have often been puzzled to see no treatment of this explicitly.

However, they are related. The existence of the secondary asset market encourages (over-) investment in production.

In my model, both are hoarding. The direct investment in production is not monolithic and needs to be broken down according to how much of the investment goes to labor. It further needs to be classified as to the society's need for the item produces. Roughly, off the top of my head:

- investment in automated production of goods not in short supply produces deflation and unemployment

- investment in production of goods (by any method) which are in short supply (for which there is demand) produces employment and raises standards of living

-- if by automation, it raises everyone's standard of living

-- if by labor, it usually lowers the laborer's standard of living

- investment in goods not in short supply produces short term deflation and unemployment; in the long term it may actually produce higher prices by shaking out competition through pricing pressure and producing monopolies

So it seems to me the total situation is very complicated. Every strategy is "good" in some circumstances and "bad" in others. It is no wonder no one has figured it out. I think the place to start is with a simple economy, e.g. why cannot the villagers of Bangladesh (or any isolated economy) feed themselves?

I have been taking a one week holiday around New Years Day. Schooling has begun and I finished the first lecture of this year an hour ago. Another one is waiting for me. I cannot spend enough time to discuss in detail.

I will discuss only two points. (1) About the relation between real economy and asset economy. (2) On your (Robert's) opinion that the total situation is very complicated.

On the 2nd point, I totally agree with you. That is why we have to make rather too simple a division. Any theoretical examinations start from this. At the same time, we have to analyze in depth the details of what are happening in the actual economy. It is quite difficult to cover two faces of necessities. We cannot go further without a useful collaboration. I am expecting that ResearchGate provides us such a possibility.

On the 1st point, I came to know that the effects of asset economy (finance economy) are very different from country to country. An article in Fuji Sankei Business-i (Jan 6 2014) by Tamura Hideo tells that there are big differences in how asset economy effects real economy in Japan and in USA. During this one year time, Tokyo stock market enjoyed high prices. For example, Nikkei index was raised from some 8,000 Yen to now 15,000 Yen. It is reported that consumption in most luxury items like jewelry and high priced watches (more than 6,000 US Dollars) are selling well. But, according to Tamura, other ordinary consumer goods are not selling well. The government is trying to promote civil (i.e. non-public) investments but it is reported that investment of big enterprises are decreasing (according either to BOJ survey and MOF legal enterprise statistics).

The situation in USA seems quite different. This difference in effects of stock market high price seems to come from the fact that US consumers possess higher rate of stocks and other risky high yield assets whereas Japanese prefer to possess their assets in the form of bank deposits.

We see how difficult it is to generalize one country's experience to other countries.

Yoshinori, the preference of Japanese investors for bank deposits, taken at face value in an international economy where the friction of owning either equities or bank deposits in other countries is quite low, is remarkable. Especially if you next tell me it has been stable over a long period of time. As I'm sure you are well aware, such a difference of preference in a low friction market suggests weakness in rational investor theory, or any homogeneous theory (i.e. that investors or consumers or any other class of market participant can be treated as an average aggregate).

The only thing that occurs to me is that while there may not be much friction of trading or ownership, perhaps there is friction of information? Actually, friction of information could explain a lot, maybe even Bangladesh. It is sort of a new popular topic of research. For example, there is some new analysis of political issue voting in economic terms, that if the cost of information, of being informed as to how candidates feel about various issues, is greater than the economic value of the issue to the voter, then the voter will prefer to remain uninformed and take blind luck. In the cases we are talking about, there are several information issues I can think of offhand, none of which are original, I have read them different places:

- Cost to citizens of one country of being informed about either the currency or market of another country enough to have confidence to invest. This cost is huge.

- Cost to investors of learning what the best investment strategy is (presumably to hold broad indices if one believes economic theory, but it took even one famous economist I know two decades to accept this in his own investing, and most people are just batted about by media experts and investment marketers)

- Difficulty to economists of assessing probability of infrequent huge crises (200 years plus) since there are few to study, but this information may culturally influence investor preferences (not necessarily rationally). (In the Roman Empire, a good survival strategy, for example, was to be a slave, as one didn't have to serve in the army and the owner encouraged reproduction. - But ultimately the owners were wiped out and the slaves remained, with their bias in favor of the working life.)

- Difficulty of knowing how to organize an economy (as in the case of Bangladesh) ... generally economists have come down on the side of Adam Smith. Even liberal economists will not tolerate too much detail in planning and organizing economies.

On the topic of consumption, one tidbit. My sister and I both consume much less than we can or than we'd like to. But it seems to us, and this is of course only an ad hoc data point, that it has become too difficult to do so in the last ten years. The rate of dissatisfaction with new purchases runs over 50% for non-routine items, meaning it has to be returned to the store, consuming a lot of time, or discarded. It seems to us that about 10 years ago the return rate was less than half that.

Hello Yoshinori Shiozawa,

I have been researching another model for Effective Demand. The results have been quite encouraging.

Here is a link to a synopsis of my work.

http://effectivedemand.typepad.com/ed/synopsis-of-the-effective-demand-research.html

This is the link to my blog.

http://effectivedemand.typepad.com/

My model basically says that the utilization of labor and capital will be capped by a measure of effective labor share. As the utilization of labor and capital goes over effective labor share, capital starts consuming its own income and stops, thus giving us the appearance of an effective demand limit.

best regards

Dear Edward Lambert ,

thank you for revitalizing this question page. I went to your blog. There are so many terms I have to learn. So, I cannot understand for the moment what you are doing. One thing I can say is that you have an idea very different from the standard models. This is a good sign for me. Let me more time until I understand roughly what you are doing.

I have posed this question more than six years ago. It may be strange to add this comment myself, but I believe that we have produced a firm theory that grasps the essence of the principle of effective demand.

Please see our book:

Book Microfoundations of Evolutionary Economics

We have written briefly why it can provide a microfoundations for Post Keynesian economics. The main reason of this claim is that we have successfully reformulated the principle of effective demand. We did not claimed that it provides the microdoundations for New Keyensian economics by two reasons:

(1) New Keynesian economics has abandoned the concept of effective demand. It keeps the notion of aggregate demand, but our new formulation re-builds the principle of effective demand at the firm or product level.

(2) New Keynesian economics admits Walrasian economics (including Arrow and Debreu 1954) as the basis of its microfoundations, whereas our theory is paradigmatically different from Walrasian economics or general equilibrium analyses. We have consciously excluded equilibrium framework anywhere in our analyses.

If some of you readers have objections, please post your opinions here. We can discuss them.