Neoclassical economics thinks in terms of a market system in which supply equals demand in every market. That includes labor markets, so unemployment could not occur in that kind of a market system; but that is an assumption, not a fact. Keynes seems to have thought in terms of a market system in which disequilibrium can occur in some markets, including labor markets, and in which disequilibrium can spread contagiously from one market to another. Keynes' idea was that, when this spreading disequilibrium settles down, there would be a kind of equilibrium -- not a supply and demand equilibrium, of course, but some other kind of equilibrium. Some later economists have suggested that "general disequilibrium" might be a better term.
Let's see how that might happen in terms of supply-and-demand diagrams. We will begin with the market for "widgets," shown in Figure 4.
In the diagram, curve MC is the marginal cost curve for widgets. We begin with equilibrium in the market for widgets with demand curve D1 at price p1. Then, for any reason, demand for widgets decreases to D2. Neoclassical economics tells us that the new equilibrium will be at price p3. Suppose, however, that prices don't drop quite that far. Instead prices only go to p2. Perhaps this occurs because businessmen don't know just how far they need to cut their prices, and are cautious to avoid cutting too much. At a price of p2, they can only sell Qd of output, as shown in Figure 5.
If the price had gone all the way down to p3, the industry would sell Qe and we would have market equilibrium in the widget market. But producing Qd at a price of p2, businessmen are not maximizing their short-run profits. Perhaps they feel they will do better in a long-run sense by limiting their price cuts. Anyway, we have "disequilibrium" in the sense that production is not on the marginal cost curve. At p2, businessmen can only sell Qd, and they will not produce more than they can sell permanently.
Here is a qualification. Businessmen might temporarily produce more than Qd, in order to build up their inventories. But there is a limit to how much inventories they want, so they will cut their production back to Qd eventually.
With a reduction of demand for widgets, any economist would expect a reduction in the quantity of widgets produced. Neoclassical economics leads us to expect that the price would drop to p3 and output cut back to Qe. At the same time, a certain number of workers would be laid off and would switch their efforts into their second-best alternatives, working in other industries, perhaps at somewhat lower wages. But the disequilibrium model says that production and layoffs would go even further, with output dropping to Qd.
Figure 6 shows the labor market. Line S is the supply of labor to the widget industry. Line DL shows the demand for labor if the widget market is at its new equilibrium, that is, if Qe is produced. However, with widget production at the disequilibrium level of Qd, demand for labor drops still further, to Dn.
With labor demand DL, the labor market equilibrium wage would be w1,
and Ne would be employed. Labor income can be visualized by the rectangle
set off by w1, Ne, the origin, and the intersection of the
supply of labor with DL. However, because of disequilibrium, demand is
less, at Dn Therefore, one of two things will happen. Either the wage
will drop still further to w2, with N1 employed, or employment
will drop still further to N2. Either way, labor income is reduced relative
to equilibrium. If the wage drops, income is the rectangle shaded like this: 
, while if the wage
remains at w1, income is the rectangle shaded by 
. Which is larger -- whether workers
are less badly off with lower wages or with more unemployment -- depends on the shape
of Dn, but either way, labor income is less than it would be at equilibrium.
Profits are likely to be down too, since businessmen don't make any profits on widgets
they don't sell, and sales of widgets are down. But these reductions in income will
lead to further decreases in the demand for widgets, as shown in Figure 7.
In Figure 7, the demand for widgets is further reduced to D3 and production of widgets to Qc. But this further reduction in production of widgets leads to a further decline in demand for labor, and that leads to a further decline in income, which leads to yet another decline in the production of widgets, and so on.
Of course, the reduction in income doesn't only reduce the demand for widgets -- it reduces the demand for all other normal goods as well. That is, disequilibrium is spread contagiously through many different goods markets, through the effect of disequilibrium on income. But this is a good news -- bad news story. The good news is that the decrease in income is spread over many different markets, so it may have little impact on the widget market in particular. This means that, if the initial reduction in demand should affect only the widget market, then the second round of reduction of demand would be very slight -- Qc would be very close to Qd. and demand could stabilize quite close to Qd. But the bad news is that, if demand reductions affect many different industries, each industry will see further decreases in demand because of the reductions in production in many other industries. The whole economy could be drawn down in a spiral of decreasing production, decreasing income and decreasing demand.
This was the sort of possibility that Keynes wanted to explore. He felt that the "spiral of decreasing production, decreasing income and decreasing demand" would eventually stop somewhere, and wanted to think of that stopping point as a kind of equilibrium. It would not be a supply-demand equilibrium, of course. Instead it would be an income-demand equilibrium.
The rest of this chapter, and the next three, will be devoted to that alternative idea of equilibrium. First, however, we need to give some time to the critics of the whole approach.
Critics of the Disequilibrium Idea
