(This is an excerpt from my book Abolish Money (From Economics)! (affiliate link), which I do not think has been previously published. In case it is not obvious, I have been tied up up with non-writing tasks in recent months. Luckily, there is a light at the end of the tunnel of distractions.)
- the sectors of the economy that define the model,
- the accounting relationships amongst those sectors, and
- the behavioural rules the sectors follow.
There is no doubt that the behavioural rules are important, as they define the operating characteristics of a model. However, we also need to make sure we properly track the accounting relationships between the sectors.
One somewhat silly example I used on my website illustrates the importance of accounting (link). I took a standard extremely simple stock flow consistent (SFC) model, and then modified the behaviour of the business sector. (Stock flow consistent models are a standard form of models used by post-Keynesians.)
In the standard version of the model, the business sector had a 0% profit margin; it hired enough workers so that the wage bill equalled the business sector revenue. I modified this behaviour to the following: the business sector always ensured that it had a 10% profit margin.
Although it sounds like an innocuous change, the model behaviour was greatly modified. Since I had not specified what the business sector was doing with its profits, it ended up accumulating an increasing stock of financial assets. In turn, this forced the government to run perpetual deficits, as it was the sole supplier of financial assets. (In the base case model, the government moved towards a balanced budget, as the stock of financial assets converged to a fixed amount.)
The model was unrealistic, but it illustrated a key point: we need to look at the entire macroeconomic system, and the linkages between sectors, in order to predict the effect of behavioural changes. There is a widespread belief that the government determines the level of the fiscal deficit. However, In this case, the perpetual deficits were the result of a change in business sector behaviour.
Financial Assets as the Glue in ModelsMainstream economic theorists want to focus on real variables: the number of widgets produced, the number of people working, etc. Financial assets are just supposed to be a “veil” over the underlying real transactions. This belief is an underlying reason why mainstream models are uniformly terrible in describing the real world.
More realistic models account for the fact that the world is uncertain; we do not know exactly how much we will earn or spend over the coming year. When outcomes deviate from plans, we need to use financial assets to buffer the uncertainty. For example, if we spend more than we expected during the month, we either have to run down our financial assets (or borrow, which is issuing a financial asset to the lender).
Within a model, the change in financial assets for a sector is equal to the sum of all of the transactions that sector has with the other sectors. Meanwhile, the breakdown of which financial assets are held depends on the model’s assumptions for portfolio weighting behaviour. For example, the household sector might allocate between zero-interest cash and interest-bearing bonds and bills based on the (real) interest rate.
Historically, economists said that money acted as the buffer stock for uncertainty. However, that is not true, other than for very short time horizons (which we cannot hope to model). Households, firms, and governments do not just adjust their money holdings in response to surprises; they adjust their entire balance sheet of financial assets and liabilities. Although the instruments in the various measures of the money supply are convenient for settlement, the big movements in balance sheets may be in long-term financial assets.
The following essay, “Money in SFC Models,” gives a more detailed analysis. (That essay was published in draft form here.)