Re: Does Debt Matter

[wfhummel@xxxxxxxxxxxxxx (William F. Hummel)]

Per wrote:
>First, the relation between private wealth and private demand should not be
>very surprising. The Friedman-Modigliani theories of consumption stress the
>importance of wealth for private consumption demand. These theories have
>been shown to yield very good predictions of consumer demand in most
>industrialised countries. Investment demand should also be heavily dependent
>on the own capital of firms. Which means that both consumption and
>investment demand are related to wealth.
>
>[I should add here that my own investigations into the Swedish case show
>that wealth works fine as the ONLY economic factor (aside age structure of
>population etc) in explaining consumption -- provided wealth is factored
>into a volume and a price-level component. Adding flow variables like
>non-property disposable income does not add any explanatory power to the
>regressions. Indeed, the coefficient for disposable income (roughly 0.3 --
>the "marginal propensity to consume") indicate that there is in fact NO
>CAUSAL RELATION going from disposable income to consumption. This is so
>because the households' wealth is roughly 3 x annual consumption, and since
>annual data has been used, the effects of wealth changes WITHIN each year
>would account for this disposable income effect.]

William:
Thanks for clarifying the point for me.  In your analysis of the
Swedish case you conclude that annual consumption is not
explained by disposable income since it can be fully explained by
the wealth effect.  That raises the question as to the relative
importance of wealth and income on consumption.  Assuming
hypothetically that income were zero, would the same annual
consumption rate exist (about 1/3 of wealth)?  I can imagine it
might for the few very wealthy, but for the vast majority in the
middle and lower income groups, it seems unlikely.  Thus in the
aggregate can we really conclude that the wealth effect is all
that matters in determining the consumption rate?
>
>Secondly, the example I set out was in terms of proportions, percentages,
>and highly simplified. When dealing with demand--supply relations we must
>add some dynamics to the scheme. But before we do, we must make clear the
>meaning of "supply". My own opinion on this is clear: The domestic product
>is basically a flow of services, and services must be sold on a fix-price
>basis, the quantity being determined by the level of demand.

William:
When you say the domestic product is basically a flow of
services, are you using the term "services" in a generic sense to
include goods and services, or do you mean specifically services
as distinct from goods?  If the latter, how do you justify
excluding goods.

>This means that
>the quantity of services EFFECTIVELY supplied must adapt passively to the
>demand. But there is another category of EQUILIBRIUM supply, which does not
>adapt to the level of demand. The equilibrium supply level is set by a host
>of factors, like the stock of real and human capital, the level and
>structure of taxation, the level of (real) rates of interest, etc. Whenever
>this equilibrium supply level is exceeded by the aggregate demand
>(=effective supply), then prices and wages will tend to inflate.

William:
Your term "equilibrium supply", which I find confusing, seems to
mean simply the supply that cannot be exceeded at any given
moment without causing an imbalance in supply and demand at the
current price level.  Is that in fact what you mean, or is there
something about supply that can reach an equilibrium independent
of demand?

>In my simplified example, I had tacitly assumed that this level of
>equilibrium supply is constant. In reality it grows with the human and real
>capital stock, and it is also affected by rates of taxation and interest. By
>lowering taxation and interest rates, the point of equilibrium can be moved
>to a higher level, allowing for more aggregate demand (= effective supply)
>without provoking inflation. The limit seems to be zero taxation and zero
>real interest rates, which, I believe, were (again) in the final arithmetic
>example I gave.

William:
This may be a partial answer to my previous question.

>William wrote:
>> I think a small positive inflation rate is desirable since it acts as a
>lubricant to the economy.  However a fluctuating inflation rate in the range
>of 2 to 8 percent would be very troublesome in investment planning, and
>would negatively impact business efficiency.  Does the historical record
>show that good long term real growth can coexist with a persistent inflation
>rate in the high end of your range? <
>
>Per:
>I don't think we have any major differences in opinion here. The 2--8
>percent range are the TOLERANCE LIMITS I would advocate. That is: inflation
>should preferrably not be allowed to exceed 8 percent, nor to go below 2
>percent at any time. The reason for these limits is that both double-digit
>inflation and near-zero inflation (not to speak of deflation proper) may be
>harmful to growth and economic stability.
>
>In order to keep within these limits, the government (not necessarily the
>central bank) should try to keep inflation somewhere near the midpoint of
>this interval. Since the variability of inflation is clearly an increasing
>function of the rate of inflation, I reckon the target should be set
>slightly below the arithmetical midpoint (which, of course, is 5%). I guess
>a 4 percent inflation target would be reasonable. I should add here that
>there may be situations when a 4 percent inflation rate is too low. As you
>know, I have been advocating a policy of keeping r - g(API) = 0. If the
>difference between the GDP deflator and the API rate grows to e.g. 3
>percent, then a 4 percent GDP deflator target would imply nominal short
>rates of r = 1 percent. Clearly r cannot go below zero, so if the GDP
>deflator--API rate difference threatens to send r below zero, the GDP
>deflator inflation target should be put somewhat higher to avoid this. I
>guess this would be a rare occasion, however.

William:
>From this analysis, it appears that you believe a fluctuating
inflation rate in the range of 2 to 8 percent is an acceptable
state of affairs.  In my opinion that is far too large a range to
be considered acceptable for efficient business planning.  How
about something like 1.5 to 3.5 percent?  During the last seven
years in the U.S. the inflation rate has stayed within a range of
about 1 percent, while the unemployment rate has steadily
dropped.
>
>William before:
>> The analysis here seems to imply that some fraction of the debt can be
>"spent."  Certainly the interest earnings can be spent, but I don't know how
>the bonds themselves can be spent.  They can be sold to other parties, but
>that merely transfers liquidity between the parties.  It does not by itself
>increase spending power.  <
>
>Per:
>Oh, you are mixing up the income--spending reflux rate with the
>wealth--spending turnover rate here. The former is really much more abstract
>than the latter, even if you seem to say the opposite here. Wealth can be
>exchanged for money (just changing the FORM of wealth), and money can be
>spent. Income, on the other hand, measures the rate of inflow of wealth
>(oftenmost in the form of money). Income as such cannot be spent, only the
>wealth which accrues by successively earning income. I prefer to think of
>income as the RESULT OF SPENDING, rather as the source of spending, if you
>see what I mean. Does this make sense??

William:
Yes, I see your point.  I got lost in the trees here.

>William wrote:
>> A related issue that you have not addressed is the private sector's
>incentive to buy government debt.  In other words, the ROI must be more
>attractive than alternatives.  I think you need to address the issue of
>interest rates on the debt and the fact that the total interest cost on the
>debt increases without limit absent taxes. <
>
>Per:
>There is no such thing as the private sector's incentive to buy gov't debt.
>The gov't needs not compete with anyone in this sense, since the gov't has a
>monopoly on fiat money (consult Warren on this one!!).

William:
It is certainly true that government has a monopoly on fiat
money.  Whether it can do anything it wishes with it is quite
another matter.  Warren's model really depends on the assumption
of coercive taxation which does not in fact exist.  We have gone
around on this issue before, and I won't belabor it here.

>Per:
>The only limit there
>is to the size of gov't debt, is the turnover of private wealth into
>spending. Gov't net debt will add to private wealth, and adding to private
>wealth will fuel private demand. Obviously there is a limit to this, when
>the sum of gov't and private demand goes up above the "equilibrium supply"
>level and starts causing serious inflation. So, inflation is the only limit
>to gov't debt.

>A Very High National Debt (VHND) will make possible, provided the economy is
>growing secularly, and that real interest rates are kept low, a quite high
>rate of gov't borrowing without affecting the debt/GDP ratio. Thus a fairly
>large public sector can be maintained without taxing a penny! The obstacle
>for a VHND regime is that the private sector's wealth-to-spending turnover
>rate is too high. Therefore, large savings incentives may prove necessary to
>make room for a VHND without fuelling over-spending and inflation.

William:
I would be interested to know what your bottom line is on VHND.
When you make it contingent on a secular growth in the economy,
are you referring to per capita growth or absolute growth?  Can
real interest rates be kept low without creating other problems?
Can you see a practical way to induce the required savings rate?
Would a mandatory purchase and ownership of bonds in proportion
to income be a viable scheme?

William F. Hummel