Re: Bush busts Social Security with massive deficits

From: Igor (jjweatherby_at_houston.rr.com)
Date: 10/03/04 

Date: Sun, 03 Oct 2004 06:20:26 GMT

William F Hummel wrote:

>>So it would hard to even say a positive correlation proves crowding out
>>unless the Fed policy and other factors that affect interest rates are
>>controlled for.
>
>
> I don't know what you mean here by "crowding out". It was not a part
> of the previous discussion.
>

Sorry I assumed you had some knowledge of principals level
macroeconomics. Crowding out refers to bonds sales used to finance
deficits pushing up interest rates. This happens because the supply of
bonds in increased. This causes bond market prices to drop which implies
interest rates rise. Essentially financing a deficit raises interest
rates. The argument is called crowding out because so economist argue
this weakens the stimulus from fiscal policy. As interest rise
investment is depressed. In other words, Government investment crowds
out private investment. This offsets the rise in consumption.

>>I am not saying in anyway the correllation proves or disproves crowding
>>out. Only to say something definite we have to control for a lot of
>>factors that are also correlated with rising debt. That is not always
>>easy. If there are ommission that are correllated with rising debt then
>>there will be bias in the estimation. It is the classic problem of a
>>regressor that is correlated with the error term. There are ways to fix
>>this but they are not always easy.
>
>
> It's easy to argue there could be a bias that could reverse the sign
> of a real correlation -- if one doesn't have to prove it.
>

Anyone with any knowledge of econometrics knows that a downward bias is
possible when right hand side regressors are correlated with the error
term. A good source that you can refer to is William Greens first year
graduate text on econometrics. Kennedy's reference is written in less
advanced language or you could look at Baltagi's panel data book. I do
not have the time to explain econometrics to you or post the proof of
how bias can exist.

The real term could be positive or 0 once the bias is taken. I do not
know the result because I have not run the regressions.

>>Again ceteris paribus. We need to know what other factors were controled
>>for. Debt rises during recessions in which times it is likely that the
>>fall in Aggregate Demand will depress the rise of inflation, assuming it
>>is a demand driven recession. So simple correlation tells us little. Any
>>regression not controling for bias tells us little.
>
>
> Exactly so. Significant inflation is uncommon during periods of
> rising debt.

That does not mean debt does not cause inflation. If deficits increase
inflation and a recession decreases inflation then the effect will be
masked by recession. If the recession is pulling down inflation rates
faster than deficits are pushing them up then inflation rates fall. This
does not in any way mean that deficits did not cause the inflation rate
to be higher than it otherwise would have been. Without deficits
inflation rates would have been even lower. If you understand what a
Calculus I class is about you can understand the argument.

   The stagflation of the 1970s was an anomaly. But your
> argument that "the Fed will have to buy bonds to keep interest rates
> down so the debt will be turned into money and cause inflation"
> _assumes_ that debt causes inflation.

No it is based on the fact that when bonds sales are increased it raises
the supply of bonds. This causes market prices of bonds to drop. This
means the yield on a bond at maturity rise or interest rate rises. To
keep the interest rates from rising the Fed must increase the money
supply. This is what causes inflation the increase in the money supply
from the Fed buying bonds not the sale of bonds itself. If the Fed does
not defend the interest rate it is not necessarily inflationary.

> Look at the 50-year history of
> rising government debt and the inflation rate which is no higher today
> than it was for most of that period.

No it is not. Over the 1990's the inflation rate dropped significantly
to around 2%. Prior to the 1990's excluding the oil shocks the rate was
averaging around 4%. The 1990's saw inflation decrease significantly due
to tight monetary policy and the Clinton tax increase.

Again this is ceteris paribus. Interest rates have been steady and have
been lowered due to Fed policy. Over the 1980's the interest rates
soared as the deficit rose. The Fed let interest rise and even
influenced to rise more inflation was in check but only because the Fed
did not monetize the debt. The sold bonds to make interest rates even
higher and keep inflation down. The high interest rates led to a strong
dollar and big trade deficits.

> There is a much easier way to avoid "insolvency" of the SS. Congress
> could declare a higher interest rate on trust fund bonds or simply
> issue a dividend to the trust fund. It's purely a bookkeeping matter,
> internal to the government.
>

Either way that means even large deficits that at some of which will
need to be paid. Raising the dividend or the interest rate does not
magically print money. Only the Fed can do that. This will raise
expenditures from the general fund and just mean we are using FICA to
finance Social Security much earlier.

The only way to avoid massive debt is raise taxes and do it early or
reduce benefits. A sensible approach would be to move back minimum
retirement ages. We are living longer and living more healthier in early
older years. People can potentially and are working longer. It makes
sense to move the minimum retirement age back. It will not solve
everything but it will help.

>>This is true but the increase in the money supply do to monetization is
>>inflationary. There is no way around it. The debts can cause inflation
>>in the future if more of the debt needs to be monetized in order to keep
>>short term interest rates in line. When the Fed monetizes debt it
>>injects money into the system. Increasing the money supply increases
>>inflation.
>
>
> No. Monetizing the debt is not done to increase the money supply.

It does not matter the purpose it is the effect. Monetizing the debt
injects money to the banking system it puts money into the system that
previously was not there.

> The net amount of debt monetized to keep the short term interest is
> nearly zero.

Are we talking about the same thing here? The principles of
macroeconomics definition of monetization of the debt is when the
government buys US treasury in order to keep the interest rate from
rising after a deficit has occurred. This is not something the Fed has
to do but it has the option to do. You are using a different definition.
That is fine if you want to define the term differently lets just
understand what concept we are talking about here.

>>It will imply increased inflation as the debt is monetized. This will
>>need to be done to keep short term interest rates in line. If the Fed
>>does abandon interest rate controls the increase in debt will increase
>>interest rates.
>
>
> This is quite wrong as explained above. Control of short term
> interest rates involves a trivial amount of debt monetization, and
> there is almost no secular component related to that purpose.
>

Really how do you think the Fed controls short term interest rates? How
do you think they reach the targets they set? They do it through open
market operations. When they increase the money supply to lower interest
rates they buy US government issued securities. They do this buying
bonds they right a check on the Fed. This is money that did not exist in
the banking system before. It is money created out of thin air to buy
this bond. It adds to the supply of money in the market. This turns the
debt bought into money. This is called monetizing the debt. Almost 100%
of Fed policy is Open Market Operations. By your definition, this may
not be monetization. So if you disagree with the use of the term that is
fine but it does not change how things operate.

>>You are increasing the supply of bonds on the market.
>>This drives down bond prices and therefore increases the interest rate.
>
>
> Be careful when you refer to "the interest rate". A generous supply
> of long term bonds will reduce the long term interest rate. However
> the interest rate on T-bills closely tracks the Fed funds rate and
> would be virtually unaffected.
>

Here you go again with false logic. You believe the Fed funds rate will
not be affected so since the interest rate on T-bills is correlated they
will not be affected. I think you need to think about causation. Say
perhaps movement in the T-Bill rate causes movement in the Fed funds
rate. Or perhaps neither causes the other. perhaps how much money is
supplied and how much money is demanded determines both rates. No it
couldn't be that could it?

>
>>The Fed can offset this but it is inflationary to do so. When ever the
>>Fed injects money it will raise inflation. Even if it replaces cash
>>withdrawals.
>
>
> But the Fed does not inject money on its own initiative as explained
> above.
>
I suggest you pick up a textbook on economics or at least look at the
Fed board of Governor's site and learn what monetary policy is and how
it is conducted. Here are some sites.

http://www.federalreserve.gov/fomc/
http://www.federalreserve.gov/policy.htm

This explains things a bit more
http://www.frbsf.org/publications/federalreserve/monetary/index.html

Once you start to understand buying bonds by the Fed is an injection of
money this begins to make sense.

>
>>This just means the money multiplier will fall short
>>because not as many new loans can be made.
>
>
> The "money multiplier" is a myth that pervades econ textbooks and
> won't die. Each generation of authors copies the last without giving
> thought to whether the concept makes sense.
>

It is not a myth. Injections multiply. No it is not exactly 1 over the
required reserve ratio. Any good principles book points out that cash
leakages stop the multiplication and reduce the multiplier. It seems
like a myth to you because you probably read a horrible textbook and
horrible principles textbooks do exist. I was forced to teach from one
for a semester.

> But it is important to note that the money supply is not under the
> control of the Fed. In fact the Fed cannot even control the size of
> the monetary base, or the component held by banks. In short, both the
> monetary base and bank credit are endogenous variables.

No it is not under complete control by the Fed but the Fed can and does
inject money into the supply and takes money out of the supply. The Fed
and banks create money. Very true the money supply is an endogenous
variable if it were not Open Market Operations could not influence it.
No the Fed can not add precisely X dollars with certainty. It can inject
Y dollars into the system and hope it becomes X dollars after multiple
deposit creation. However, MDC is not guaranteed. People can hold money
in cash at any point in the line stopping the process cold. So how much
money is created when the Fed buys one dollars worth on bonds depends on
if people deposit the proceeds in the bank or if they hold it in cash. I
never said the Fed forces people to keep money in the bank or the bank
to make a loan. Yet they do add money to the system. The total amount is
influenced by how much the banks lend and how much people withdraw and
hold in cash.

Relevant Pages
Loading