In a previous post I wrote about the key reason that the natural rate of interest does not exist. There I discussed the Kahn Multiplier. We saw that when investment increased consumption increased along with it due to the fresh income received from the investment spending. The increase in consumption was then multiplied out into the wider economy in accordance with the Kahn Multiplier. This increase in consumption must be met by productive capacity that is already in existence. In this post, I want to get away from the fantasy constructions that marginalists use and explain how the multiplier functions in the real world.
A nice way to introduce this might be to tell a story from my own personal experience. About a year and a half ago I was at a party with an old school friend. He is an engineer and he was at the time working as a supervisor on a production line at a pharmaceutical company. He, like many engineers and practical intelligent people, is quite skeptical about economics and economists. I told him that he would get a real laugh out of the economic theory of the firm.
When I described the theory of rising marginal costs and marginal revenue he laughed. “Are you serious?” he said, “Do economists honestly think that we could run a plant on that basis? That is completely absurd.” I said that I knew it was and then laid out the Post-Keynesian theory of the firm. Then he nodded and said “Yes, that is basically how it works. I’m glad some of you economists have a clue what you’re talking about.”
So, what is the Post-Keynesian theory of the firm? Well, Lord Keynes has laid it out in detail in an excellent post over at his blog entitled Price, Average Total Cost, Average Variable Cost and Marginal Cost. If you want a comprehensive exposition with full references I suggest checking this post out. Here I will just go over the bare bones.
The Post-Keynesian theory of the firm is best laid out by the following diagram.
So, what does this mean. Well, the UC-curve stands for ‘unit costs’. These fall over time due to increased economies of scale. That is, until the reach the point FC. FC stands for ‘full capacity’. I will describe this in a moment. We should note beforehand that the MC-curve, that is, marginal costs, is flat up to the point FC.
So, what is this point FC? That is full capacity. At this point costs will begin to rise. This is because workers will have to be paid overtime and machines will be operated for longer than they should be and will not get sufficient repair. The point FCth is theoretical full capacity. This is the point at which a plant will be literally pushed to the limit. I.e. when machines are literally operating at full capacity and no more units could possibly be cranked out of them.
Now, the important thing is that firms will typically be operating at the point where the MC-curve is flat and the UC-curve declining. This is because they build excess capacity into the plant. A typical plant, during an economic upswing will be operating somewhere around 90% of capacity. (My engineer friend reported that in the wake of the serious recession in Ireland his plant was operating below 80% of capacity which was exceptionally low).
The engineers build this excess capacity in to keep slack in the system. They know that the real economic world is tumultuous and so they have many ‘buffers’ in place in the production process to ensure that everything goes smoothly. Their nightmare is a plant operating at the point FC. For them to even consider their plant operating at the point FCth is, although an interesting thought experiment, a little bit ridiculous.
Anyway, back to the multiplier. When investment increases and consumption rises in line with the multiplier what happens is that some of this slack is picked up in the production process. Typically this will not lead to increased costs — indeed, it may even lead to decreased costs as economies of scale kick in. But during a major boom we might see plants reach the point FC for a short period of time. The main thing causing the rise in costs and hence in prices here will increased overtime payments.
This is, of course, why Post-Keynesians stress the wage-led nature of inflation over the demand-led inflation that many mainstream economists instinctively think about. When demand rises past a certain point, it is wages that rise first. These are then passed on to consumers in the form of higher prices. The process is not one where consumers go into supermarkets and start using their larger pay packets to bid up prices. That is ridiculous and only exists as an oddity in the minds of economists who never got away from Walras and his auctioneer.
So, that is how the multiplier works in the real world. When investment increases the inbuilt slack in the production process allows for the increased consumption without price inflation. This is not just in manufacturing plants either. Even barbers have inbuilt slack in the form of extra chairs in their barber shops. Again, prices will only start to rise if overtime is being paid for an extended period. The vast majority of the time, however, the economy is not up to the point of full capacity and wages will only rise in line with productivity growth. (Indeed, since the 1970s they haven’t even kept up with productivity growth!).
One more point. If the economy were kept at high full employment — I would estimate that this would be around 2-3% unemployment — for a very long period of time, maybe 5-10 years and this looked like it might be a permanent state of affairs, then producers would increase the level of excess capacity and hence slack in the production process. A new norm would be built into the production process. There is a chance that the level of high full employment could then be increased. We have never seen such an experiment tried before, however. Any time the economy has been pushed this hard — I think of wartime — wage and price controls were in place to manage demand directly. So, we do not know if such an effect would ever occur in the real world. It would be interesting to study, however. But these days we can barely even dream of even low full employment.
Nice post, Phil. Again, economists seem to have a problem with cause and effect. BTW, here where I live in Maine we had an example of a very nice miik producer going under because of failure to have excess capacity. You can read about it here.
Great example. Firms in the real world know that they face a dual problem. One is to get more demand for their product by increasing market share. This is the job of the marketing department. The other is to ensure that they have adequate and flexible supply in order to meet this new demand. This is the job of the engineers.
When viewed from this perspective the marginalist story is truly absurd.
“Even barbers have inbuilt slack in the form of extra chairs in their barber shops. Again, prices will only start to rise if overtime is being paid for an extended period.”
Barbers are an interesting one, which I’ve observed closely over the years (being a man and all that).
Generally here the chairs are rented at fixed cost and the barbers are actually self-employed. So what happens is that time at work just expands and the barbers work longer and longer hours (and try to go faster) while earning at the same rate (because they own the profit share above the fixed cost of the chair). So the costs never rise.
Near perfect frictionless dynamic expansion.
I think you could get increases in potential output if economy without higher investment also. People can innovate and become more resourceful. They can get more out of the same capital over time.
Therefore you don’t always need to increase investment to support higher consumption. FC varies even if investment is the same.
In first sentence above I meant “in economy” not “if economy”.
Yes. There is some research into efficiency wages that suggests people will be more productive if, for example, they are given higher wages. Obviously it is also becoming increasingly well-known that different management techniques promote different levels of productivity and research is coming increasingly to show that more “hands off” approaches to management increase productivity, at least in the creative industries.
One thing which happens when unemployment is low and demand increases is that companies try to buy people away from other companies. Last time this really happened in the Netherlands was during the dotcom bubble. This, of course, leads to quite rapid wage increases (which to an extent may be efficient, when productive companies are able to pay the highest wages).
Good point! Yes, this happened in Australia too until recently.
The CFO and the firm’s bank, board and investors view it differently from the view of the factory floor guy. They run a CAPM model. CAPM inherently assumes loanable funds theory. When rates fall because of increased saving, baseline sales fall and projected sales are based on a real baseline. Without money-printing, it is more likely that producers’/investors’ cost projections are in line with reality. The fall in WACC is based on real savings the purpose of which is to maintain/increase consumption in the future. The cost to produce hats is reduced by virtue of an increase in the number of people saving to buy hats. The producer/investor see the price and may not be aware of why it changed – they don’t have to be as long as the price is real.
When the central bank artificially reduces rates, WACC falls while consumers save less and spend more, thus increasing baseline sales. With increased consumer access to credit, baseline sales rise even more as people buy hats on their cards and pay only the minimum. The more credit expansion that takes place, the more future cost inflation is baked into the cake. A few years in, WACC has recovered, people who spent their savings cannot increase their purchases and people who went into debt to finance their former level of consumption have to cut back. Meanwhile production costs have gone up. Both consumers and producers are overextended, and mal-investments are revealed. In the aftermath, you’re darned right that spooked investors’ liquidity preference has increased. That’s not the market’s fault, nor is it the market’s fault that demand isn’t what you projected – – “insufficient aggregate demand” means that the problem is that reality didn’t meet your rose-colored projections – – the Fed put those rose-colored glasses on you – – blame them, not reality.