The Meaning of Supply and Demand

Fred Foldvary

[Reprinted from a Land-Theory online discussion, December 1999]

Roger Sandilands:

But if we are interested in the underlying nature of the real economy, and want to get behind the "money veil", then a shift in the demand schedule is the same thing as a shift in the supply schedule. If tomorrow everyone were 5 percent more productive then (abstracting from differing demand elasticities, or assuming that resources shift easily from low to high demand elasticity sectors) the aggregate demand would also shift up by 5 percent in real terms.

Fred Foldvary:

We have more of a disagreement on terminology than on the economics. But the terminology is important in keeping our thoughts and communications clear.

The aggregate supply curve would shift out to the right to a higher output.

The aggregate demand curve depends on money income, and would stay put, since the quantity of money is the same. The greater supply would then cross the demand curve at a lower price level. If the money supply does not increase, then the effect of greater productivity is to lower prices.

If you are saying that when aggregate supply shifts out, the quantity demanded will equal the quantity supplied, I agree. But that extra quantity demanded is not due to a shift in demand but to a movement down the aggregate demand curve to a lower price level. Of course if the money supply increases by the same proportion as the supply, then the demand will shift out. But in the above, you say nothing about increasing the quantity of money flow (MV).

Roger Sandilands:

In a very flexible and competitive economy, prices would tend to fall throughout the economy by 5 percent if the money supply was held constant. Or the authorities could expand the money supply by 5 percent to keep the general price level steady.

Fred Foldvary:

If prices fall with constant money, the demand, meaning the whole curve, is not shifting. If money expands, that shifts the curve out, since for any given amount of output, the price level would be higher.

Roger Sandilands:

Now go back to microeconomic supply and demand. Why would demand for widgets increase? Maybe because widgets are the flavour of the month. But more likely because national income has increased, either in money or in real terms. If in money terms only (i.e., inflation), the costs of widget-making will almost certainly have increased too. Then the supply (cost) curve will not be independent of the shift in the demand curve. Both shift upwards from a common cause (inflation).

Fred Foldvary:

Not so. The supply curve for widgets is independent of the demand because the supply curve is a relationship between price and quantity, holding everything else constant. If the money supply increases, then everything else is not constant, and the curve shifts. The shifting of supply is not independent of demand, but the supply curve, which is only a relation between price and quantity with all else constant, is independent of everything other than price and quantity.

Roger Sandilands:

Adam Smith introduced the Wealth of Nations by stating that productivity depends on the division of labour and specialisation. And that the division of labour is limited by the size of the market. So when, in the aggregate, market demand (in real terms) increases, we can expect that outwardly shifting supply curves are the cause and the consequence. Growth has an underlying tendency to be cumulative and self-sustaining. Supply creates its own demand.

Fred Foldvary:

The last statement was by J. B. Say in the early 1800s, but he did not express it that way. "Supply creates its own demand" is from Keynes. What Say said in "Say's Law" is that in creating a supply, the factors are paid that amount that enables them to buy the supply, hence additions to quantity supplied will equal additions to quantities demanded.

Roger Sandilands:

And so, to return to Victor's original point, if the aggregate demand for loans shifts, we can be pretty sure that that shift in demand will not have been independent of the supply of loanable funds (and vice versa).

Fred Foldvary:

A shift in the demand for loans is indeed independent of the supply. The supply of funds comes from savings. The demand comes from investors and consumers. It is possible that more folks want to borrow, while folks don't want to increase savings as much. Then the interest rate rises to equilibrate the added demand.

If aggregate output is expanding while the supply and demand ratios of savings to consumption is constant, then the shift in demand for loans is matched by a shift in savings, because both are growing at the same rate. But this is not necessarily or even usually the case.

Roger Sandilands:

But Fred, what happens if real demand increases in the schedule sense?

Fred Foldvary

If demand is measured in terms of money (MV; money stock times velocity), then there is no real demand but only a nominal demand. In a moneyed economy, MV=PT (price level times transactions), and the aggregate demand is MV/P at various levels of P, holding MV constant, hence nominal (moneyed) since it is a function of M.

In a barter economy, people purchase goods with goods, so the aggregate demand is vertical and coincides with the aggregate supply, since the price level is irrelevant. The ratio of trading of one commodity against all others is irrelevant to the total demand for goods, since what is exchanged is not that one commodity but all commodities against all commodities. But then the whole concept of aggregate demand and supply is then meaningless, since these are schedules of the price level and output. There is only output, and no price level. So in effect, there is no supply and demand at all, but only some aggregate quantity of goods supplied and some aggregate quantity demanded, and they are equal.

Hence, as I see it, the concept of aggregate supply and demand imply the existence of money and a price level, and aggregate demand is not real but always nominal. The aggregate quantity of goods demanded is real, and equal to the aggregate quantity supplied, but it is not demand but a point on the demand curve, or specific item in the schedule.

Roger Sandilands:

Can you then say that this new demand schedule intersects the (vertical) supply schedule at a higher price?

Fred Foldvary:

No, because there is no real aggregate demand curve.

Roger Sandilands:

Or would you have to admit that an outward shift in the demand schedule cannot occur except via an outward shift in the supply schedule? I am genuinely puzzled by my own question.

Fred Foldvary:

I think the puzzle is resolved by realizing that aggregate demand is moneyed.

Roger Sandilands:

Perhaps the answer is that the demand schedule in real terms cannot shift outwards; that there can only be a movement down an aggregate demand schedule to effect the secular increase in *quantity demanded*?

Fred Foldvary:

That's it.

Roger Sandilands:

Another way of looking at it is to say that through time the downward-sloping real demand schedule actually coincides with a downward-sloping aggregate supply schedule.

Fred Foldvary:

No, the aggregate supply schedule is either vertical (independent of the price level) or else, in a Keynesian situation with unemployment and a fixed nominal wage, upward sloping. Over time, as output grows, the aggregate supply shifts out to the right.

\

Roger Sandilands:

But I can see some advantages to your preference for a constantly outward shifting vertical supply schedule.

Fred Foldvary:

Yes, vertical in the classical world of flexible wages and prices.

Roger Sandilands:

"In the aggregate supply is demand is supply." Why? Because we live in an exchange economy and we can only demand real things if we offer something real in exchange.

We can analyse widgets on the assumption that suppy and demand are independent, but we this is not how we should analyse the supply and demand of GDP (aggregate supply and demand).

Fred's analysis of aggregate supply and demand to determine the aggregate price level is not wrong, but it is not quite the same kind of point I was making in response to Victor's analysis of the market for loanable funds.

When we look at the supply and demand curves for widgets to determine the equilibrium price of widgets, we are looking at a particular price and assume all other prices are constant. We in effect find the *relative* price of widgets. But when Fred looks at an aggregate supply curve -- more or less vertical at full employment -- and superimposes an aggregate, downward-sloping demand curve, he is trying to find the *general* price level.

If demand shifts upward it's presumably because there has been an increase in the money supply -- an increase in *monetary* demand -- and the price level drifts upwards, more or less pari passu with the money supply. The price of widgets and everything else tends to rise in the same proportion. Nothing "real" has changed -- price relativities are unchanged. This is the neutrality-of-money proposition. Of course it's a simplification, and real things do get upset by monetary shocks in the short run.

But if we are interested in the underlying nature of the real economy, and want to get behind the "money veil", then a shift in the demand schedule is the same thing as a shift in the supply schedule. If tomorrow everyone were 5 percent more productive then (abstracting from differing demand elasticities, or assuming that resources shift easily from low to high demand elasticity sectors) the aggregate demand would also shift up by 5 percent in real terms. In a very flexible and competitive economy, prices would tend to fall throughout the economy by 5 percent if the money supply was held constant. Or the authorities could expand the money supply by 5 percent to keep the general price level steady.

Now go back to microeconomic supply and demand. Why would demand for widgets increase? Maybe because widgets are the flavour of the month. But more likely because national income has increased, either in money or in real terms. If in money terms only (i.e., inflation), the costs of widget-making will almost certainly have increased too. Then the supply (cost) curve will not be independent of the shift in the demand curve. Both shift upwards from a common cause (inflation).

If there has been a real increase in GDP (and assume no change in the overall price level) then it is because the various sectors (widgets, food, transport, etc, etc) are more productive; their unit costs have fallen and so their supply schedules shift out and down. Perhaps widget makers do not share in the productivity improvements that have caused GDP to increase. In that case their cost schedule is unchanged and the increase in demand for widgets will increase their absolute and relative price. But this must have been an exception, otherwise GDP would not have increased. Overall, individual demand schedules have shifted up because a host of supply schedules have shifted down. Overall, on the reasonable assumption that wants are insatiable, and that the *overall* income elasticity of demand is equal to unity, demand has increased in line with supply.

We could of course object that the above story ignores the business cycle, with Austrian, monetary, Georgist, Schumpeterian, or whatever causes. But in the long run, the above shows why supply is demand is supply.

Adam Smith introduced the Wealth of Nations by stating that productivity depends on the division of labour and specialisation. And that the division of labour is limited by the size of the market. So when, in the aggregate, market demand (in real terms) increases, we can expect that outwardly shifting supply curves are the cause and the consequence. Growth has an underlying tendency to be cumulative and self-sustaining. Supply creates its own demand.

Keynes was supposed to have buried Say's Law. But he only showed why monetary disturbances can interrupt the underlying trend. Henry George gave an equally plausible explanation for the interruptions, and also concentrated on the question whether the fruits of secular progress were properly and healthily distributed. But in answering the Big questions he did not play around with microeconomic supply and demand curves. For, in the aggregate, supply is demand is supply.

And so, to return to Victor's original point, if the aggregate demand for loans shifts, we can be pretty sure that that shift in demand will not have been independent of the supply of loanable funds (and vice versa).

Roger Sandilands:

This supply and demand schedule game is the stuff of introductory textbook economics. Later on, some students maybe, just maybe, learn that when one of the schedules shift it is unlikely that the other schedule will be unaffected. In the aggregate, supply is demand is supply.

Cliff Cobb:

I don't understand what you are saying here. Isn't it axiomatic that supply and demand are determined independently? I think there may be problems with that assumption, but I understood that that was the basis of all micro analysis: two equations and two unknowns.

Fred Foldvary:

We first need to clearly distinguish between the demand for a good or factor, and the aggregate macroeconomic demand for all goods. For goods, the vertical axis is price. For the economy, the vertical axis is the price level, such as measured by a price index.

The agg demand curve presumes all is held constant other than the price level and total output. So money is also constant. Given constant money, hence a constant income, the lower the price level, the more goods that amount of money can buy. Hence the agg demand slopes down.

Meanwhile, the agg supply can be vertical or slope up. So they are independent and will cross to determine the price level and total output. In a full-employment economy, the agg supply will tend to be vertical, i.e. output is independent of the price level. So an increase in money shifts out demand but only raises prices, not output.

Roger Sandilands:

But, re the implications for microeconomics, would we agree that the upshot is that while (i) there are analytic advantages in drawing individual (microeconomic) supply and demand schedules as independent of each other, with one capable of shifting without the other shifting also, nevertheless (ii) the ceteris paribus assumption that lies behind these abstractions can be highly misleading. Why? Because the forces that cause the one to shift (in particular, the size of the aggregate market) almost always are acting simultaneously on the other schedule too.

And this is likely to be particularly true of the market for loanable funds where the demand and supply schedules both depend on aggregate income. (And/or, in the market for loanable funds to finance real estate, by the speculative rise in the price of the assets that are changing hands -- so that my purchase equals your sale proceeds that you may put back into the loanable funds market. And vice versa when the bubble bursts. When the market is rising/falling, do not both the supply and demand schedules for loans increase/decrease?)