Recent Posts

Sunday, September 21, 2014

Demand Creates Its Own Supply: Government Bond Edition

Keynes summarised Say's Law as "supply creates its own demand". That is a subject that is somewhat controversial, but I do not want to discuss that here. Rather, I am interested in what might be called the Anti-Say's Law: demand creates its own supply. Although it sounds fairly silly when applied to a market like pork bellies, it actually appears applicable to the government bond market. I have touched upon this topic in earlier articles, but I want to tie it into to my recent primer on stock-flow norms. This concept provides an alternative means of understanding current trends in the developed economies.

Passive Fiscal Policy

I am in the camp that argues that the business cycle has been largely tamed by passive fiscal policy, although sufficiently large malinvestment can inject some serious volatility into the system. However, this stability implies a sluggish recovery (you typically need to introduce some instability into a system in order to get a faster reaction time). This view is not the consensus view, which is that the business cycle is dominated by the actions of monetary policy (with the disclaimer that the "zero bound" messes everything up).

To summarise, this passive policy is the result of two main effects:

  1. Most tax revenue is generated as the result of taxing economic activity; income taxes and sales taxes rise along with nominal incomes. There are some taxes proportional to real activity (fixed gasoline taxes, and tolls). In North America, the only taxes of note that are not driven by activity are property taxes, which are collected by non-central governments. This leads to local governments actually acting a pro-cyclical fashion, but this is overridden by the other components of passive fiscal policy.
  2. Government programme spending - things like defence, pay for bureaucrats - are essentially fixed in nominal terms for short periods of time. Social welfare spending tends to rise as activity falls (unemployment insurance, etc.).
Taken together, this means that budget balances are counter-cyclical, and that governments have no control over their budget balance. The latter point will come as a surprise to those who believe media discussion of the topic, possibly including bureaucrats who work in finance ministries. Finance ministries churn out detailed spending plans, and the media discusses the dollar amounts involved as if they actually mean something. Commentators commend governments for their brave stances to reduce the deficit (which they invariably announce that they will do).

The reality is that those budget numbers are entirely hypothetical. Although programme spending numbers may actually match reality, revenue and social welfare spending are entirely dependent upon the cycle. In most years, it is safe to project a small steady growth, so the forecasts can be roughly correct. In fact, governments typically have smaller deficits than forecast, as politicians, just like corporate managers, have perfected the art of "big bath accounting". 

Properly understood, fiscal policy consists of setting a few parameters - programme spending, tax rates, social welfare scales - and the budget outcome is the result of the operation of the economic system. (In economics lingo, the budget balance is endogenous.)  This means that government finances will adjust to the pressures which are created by private sector stock-flow norms.

A Recession In A Simple SFC Model

I developed a very simple Stock-Flow Consistent (SFC) model which simulated the impact of a hard debt limit on an economy, which I discussed here (spoiler: it's pretty ugly).  I discussed some further details of the model in a follow-up article, which covered some of the ground that I am now discussing. I will now explain how those results tied into the discussion of stock-flow norms.

Chart: Cash Holdings In Simulated Economy

The recession in the simulated economy (which is a no-growth economy; in steady state, all variables are constant) was the result of a rise in savings in the household sector. This is as shown above; the shock which launches the recession coincides with the rise in household cash balances. (Note that the only financial asset in this model is cash, which are the government liabilities.) The corporate sector cash balance initially falls, but it returns to a previous target level. (Note that the simulation runs above are for the case where the debt ceiling is not in effect,)

The chart above shows the savings and investment flows. In this simple economy, the only investment is inventory investment. The rise in household savings initially finances a rise in inventories, but what we see in further periods is that there is dis-investment in inventories. The chart below explains why.

Chart: Inventory-Sales Ratio In Simulated Economy

In the model, the business sector has a stock-flow norm: it wants to hold 1 period's sales volumes in inventory. At the beginning of the simulation, the amount of production equalled sales, and so inventories were also constant. The rise in savings leads to lower sales volumes, which raises the inventory-sales ratios. Production is cut back in order to get the inventory-sales ratio back towards 1. (The line in red shows that the operation of the debt ceiling prevents this.)

Chart: Simulated Industrial Production And Unemployment Rate
The chart above shows what all this means for activity -there is a recession. Unemployment rises as production is cut back. Eventually, the economy returns to full employment (assuming a debt limit is not in place).

The simple interpretation of what the model is showing is this: the demand to hold government liabilities rose, and so the supply has to be created. When you look at my earlier description of fiscal policy, the only way that the supply can appear is for there to be an (unplanned) slowdown in economic activity. (Note that in Modern Monetary Theory - MMT - the analysis of the demand for "net financial assets" is often exactly on these lines.)

I accept that this model is simple, and I discuss some of the arguments against it below. However, my feeling is that the basic principle is sound, and it explains a lot of the woes facing the developed economies today. There are forces driving for an increased accumulation of government liabilities (debt and base money) - that is, aggregate stock-flow norm parameters are changing. And so growth has to slow to create the required deficits. Correspondingly, I see austerity policies as inherently doomed, as they are leaning against well-entrenched economic forces.

Limitations Of The Debt Ceiling Model

The basic premises of my debt ceiling model could possibly enrage some supporters of free market capitalism, as it contradicts the assumption that increased savings leads to increased economic growth. (I wrote about this in my earlier articles, as well as in Savings Equals Investment, And All That, where I responded to one debate about the S=I identity.)

If we go beyond slogans, there are a number of reasonable objections to the model results I present. As I explain, I do not find them insurmountable.

The most important objection is that this simplified model does not allow for private sector emission of debts (there is the equity of the business sector). This is a limitation of the model, and is not a feature of more sophisticated Stock-Flow Consistent models. My excuse is that I want to focus on fiscal policy, and leave private sector debt emission steady (an "all else equal" condition). If private sector debt levels are constant, it does not make much of a difference for cyclical analysis what level they are constant at.

More realistically, the cycle is dominated by private financial flows, which are disrupted in a recession. My simulation results would be dominated by whether or not the economy is tipped into a recession. It would be possible to show all kinds of strange effects if the initial conditions had the economy perched on the edge of recession. The inherent stability of the "debt ceiling" model limits those effects.

But yes, rising private debt allows for falling government debt. We saw this in various housing bubbles in the English-speaking world; once the supply of private debt dries up as the result of the lack of sustainability of debt loads, government debt was forced into existence the hard way.

The second effect revolves around "real balances" - the price level should fall, so as to allow the real value of money balances to rise to the higher desired level. This would be theoretically possible in my simulated economy, as there is no private debtors that will be compromised by a fall in nominal incomes. (Many members of the middle class would be crushed by their mortgage debt if nominal incomes dropped by 10%, and the loan balances were not adjusted,) In practice, we do not see anything resembling this behaviour, so I do not feel bad that my model ignores it. The weakness of the real balance effect is well known; for example see this quote of James Tobin (via Fictional Reserve Barking).

Another method to avoid a recession in this example would be for fixed investment to ramp up, matching the rise in savings. This would presumably be triggered by falling interest rates, which is why monetary policy is presumed to work. But it is unclear why businesses would ramp up investment when demand is falling. Recessions are pretty much defined by falling investment activity, and investment only recovers once the weakest firms have been eliminated and large fiscal deficits put a floor under demand.

Although these theoretical complaints strike me as weak, there is an very good practical objection to it. The business cycle is driven by fixed investment, and savings behaviour is slower moving. Changing stock-flow norms are not the prime mover behind recessions, rather they help regulate the speed of the economy during its expansionary phase. And in the current circumstances, this factor helps explain the tepid pace of growth.

(c) Brian Romanchuk 2014


  1. Brian, very nice post. And the visuals are excellent.

  2. Pretty much as I see it as well.

    The spending injections create an expansion and that creates an increased demand for risk free saving.

    I liken it to breathing. It is much easier to breathe out if you breathed in first. The more you breathe in the stronger the breath out, and the limit is your lung capacity and rib cage.

    You have to let the economy breathe.

  3. Along with the others, pretty much as I see it.

    The sequence seems (to me) to have consequences for how we view money supply and wealth. When government, seemingly operating on automatic, has a revenue shortfall, it borrows to make up the difference. If the borrowing is from banks (including the central bank), the apparent money supply in deposits at banks increases. If the borrowing is from the private sector, only the wealth of the private sector increases.

    This scenario is a description of increasing financial wealth for the economy as a whole.

    Now the part I have trouble with is the distribution of the counter-flow of real labor and resources that the financial wealth represents. Clearly, we have described an increase in financial wealth. Have we also described an increase in real physical wealth?

    I think not (despite the fact that I have no scenario that describes a lack of real physical wealth increase).


Note: Posts may be moderated, and there may be a considerable delay before they appear.

Although I welcome people who disagree with me, please be civil.

Please note that my spam comment filter appears to dislike long "anonymous" posts. I get no warning about this, and only go through my "spambox" infrequently. The best bet it to keep comments short, and if you think the spam filter struck, let me know with a short comment.