On Fed Control of the Money Supply

17 views
Skip to first unread message

William F Hummel

unread,
Mar 21, 2002, 5:03:31 PM3/21/02
to
Following is a copy of the first half of a letter Donald Kohn,
Director of the Division of Monetary Affairs under the Board of
Governors of the Federal Reserve System, wrote to Warren Mosler
in 1994:

Dear Mr. Mosler,

Chairman Greenspan asked me to respond to your recent letter
regarding your treatise "Soft Currency Economics."

You made many interesting points in the treatise. I agree in
particular that the textbook money multiplier model is an
oversimplification of the money supply process. As you point
out, the lag even in the current system of "contemporaneous"
reserve requirements implies that monetary policy cannot achieve
perfect short-run control over money or reserves. Indeed, as you
stress, in the short run the causation runs more from money (see
note below) to reserves than vice versa. However, the Federal
Reserve can control the composition of reserves between
non-borrowed and borrowed reserves fairly closely in the short
run. By varying this mix the Federal Reserve can influence the
federal funds rate and the other money market rates in the short
run and thereby affect growth of the money supply over a longer
period.

---------
"Money" means the total of checkable deposits and cash.

Experience in the 1980s demonstrated that Fed's targeting of the
money supply resulted in unacceptable volatility in the Fed funds
rate, without actually succeeding in controlling the growth rate
within rather broad targets.

William F Hummel

unread,
Mar 21, 2002, 5:06:12 PM3/21/02
to
Following is a copy of the first half of a letter Donald Kohn,
Director of the Division of Monetary Affairs under the Board of
Governors of the Federal Reserve System, wrote to Warren Mosler
in 1994:

Dear Mr. Mosler,

Chairman Greenspan asked me to respond to your recent letter
regarding your treatise "Soft Currency Economics."

You made many interesting points in the treatise. I agree in
particular that the textbook money multiplier model is an
oversimplification of the money supply process. As you point
out, the lag even in the current system of "contemporaneous"
reserve requirements implies that monetary policy cannot achieve
perfect short-run control over money or reserves. Indeed, as you
stress, in the short run the causation runs more from money (see
note below) to reserves than vice versa. However, the Federal
Reserve can control the composition of reserves between
non-borrowed and borrowed reserves fairly closely in the short
run. By varying this mix the Federal Reserve can influence the
federal funds rate and the other money market rates in the short
run and thereby affect growth of the money supply over a longer
period.

End quote


---------
"Money" means the total of checkable deposits and cash.

Experience in the 1980s demonstrated that Fed's targeting of the
money supply resulted in unacceptable volatility in the Fed funds
rate, without actually succeeding in controlling the growth rate

within what were rather broad targets.

Mark Patrick Witte

unread,
Mar 21, 2002, 10:39:30 PM3/21/02
to
Astonishing revelations! Textbooks simplify, there is not perfect
short run control of the monetary aggregates (or FFR either), the Fed can
control the money the money supply over longer periods. Wow, this guy was a
real whistle-blower!

In article <hemk9uk3lotlv18ig...@4ax.com>,

But it did succeed in its real goal, which was bringing down
inflation. It (thankfully) abandoned (partial) reserve targeting
and chose to go back rate targeting, since it couldn't do both, since they
are co-determined.

Mason Clark

unread,
Mar 21, 2002, 11:39:28 PM3/21/02
to
On 22 Mar 2002 03:39:30 GMT, mwi...@merle.acns.nwu.edu (Mark Patrick Witte) wrote:

> Astonishing revelations! Textbooks simplify, there is not perfect
>short run control of the monetary aggregates (or FFR either), the Fed can
>control the money the money supply over longer periods. Wow, this guy was a
>real whistle-blower!
>

Warren Mosler is much more than a "whistle-blower."
Mark Witte would do better to look before blowing.

An essay "Soft Currency Economics" is at:

http://www.warrenmosler.com/docs/docs/soft0004.htm

on Mosler's site:

http://www.warrenmosler.com/

While you're browsing, take a look at the Mosler MT900r:

http://www.fast-autos.net/mosler/moslermt900r.html

I'm not in full agreement with all of Warren Mosler's ideas
about economics but he poses a good question in his
Soft Currency Economics. (and I'd like to have had his
*real* knowledge of economics)

Mason C

William F Hummel

unread,
Mar 22, 2002, 10:18:16 AM3/22/02
to
On 22 Mar 2002 03:39:30 GMT, mwi...@merle.acns.nwu.edu (Mark
Patrick Witte) wrote:

> Astonishing revelations! Textbooks simplify, there is not perfect
>short run control of the monetary aggregates (or FFR either), the Fed can
>control the money the money supply over longer periods. Wow, this guy was a
>real whistle-blower!
>

Mr. Witte apparently isn't even embarrassed by ignoring the key
statement made in the letter, namely "in the short run the
causation runs more from money to reserves than vice versa",
which was the point at issue. Instead Mr Witte attempts to argue
his case, as he often does, by the use of ridicule and diversion.
In this instance he selects comments of secondary importance in
the letter to focus the readers attention away from a position he
cannot defend.

First, the Fed did not target reserves per se during the period
in question. It targeted the money aggregate M1, using its
control of reserves as the tool. Second, the inflation was
killed at the expense of double digit unemployment, at best a
mixed success. And finally the term "co-determined" is simply
economist jargon that says prices and quantify are functionally
related without saying anything about which is the control
variable and which is the independent variable. Mr. Witte has
been using the term as a smoke screen in an attempt to hide the
point at issue.

WFH

Mark Patrick Witte

unread,
Mar 22, 2002, 11:08:13 AM3/22/02
to
In article <rfcl9u0ppi6ka9bjl...@4ax.com>,

Mason Clark <mas...@ix.netcom.com> wrote:
>On 22 Mar 2002 03:39:30 GMT, mwi...@merle.acns.nwu.edu (Mark Patrick Witte) wrote:
>
>> Astonishing revelations! Textbooks simplify, there is not perfect
>>short run control of the monetary aggregates (or FFR either), the Fed can
>>control the money the money supply over longer periods. Wow, this guy was a
>>real whistle-blower!
>>
>Warren Mosler is much more than a "whistle-blower."
>Mark Witte would do better to look before blowing.

Warren did not write that letter.

Mark Patrick Witte

unread,
Mar 22, 2002, 11:53:21 AM3/22/02
to
In article <qihm9ugbt6ra4g1eh...@4ax.com>,

William F Hummel <wfhu...@attbi.com> wrote:
>On 22 Mar 2002 03:39:30 GMT, mwi...@merle.acns.nwu.edu (Mark
>Patrick Witte) wrote:
>
>> Astonishing revelations! Textbooks simplify, there is not perfect
>>short run control of the monetary aggregates (or FFR either), the Fed can
>>control the money the money supply over longer periods. Wow, this guy was a
>>real whistle-blower!
>>
>Mr. Witte apparently isn't even embarrassed by ignoring the key
>statement made in the letter, namely "in the short run the
>causation runs more from money to reserves than vice versa",
>which was the point at issue.

Ah, I would like to thank Mr. Hummel for letting me know which of
Donald Kohn's words he considered the key ones. Since I'm not him, and lack
mindreading ability, I would not have known. In fact, to me, it looked like
a the sort of polite letter I often wrote in my days working in Washington
to constituents who wrote in with empassioned but not entirely worked out
thoughts about government policy. (I know such people who have this as part
of their duties at the Fed as well.) Without Mr. Hummel's insight, I would
have thought the letter was doing the classic, "But of course what you say
is right (to some degree), *however* the more important issue is that..."
In this case, the "however" falls right after the line cited by Mr. Hummel:

>>>However, the Federal Reserve can control the composition of
>>>reserves between non-borrowed and borrowed reserves fairly
>>>closely in the short run.

From a technical standpoint, there are supply and demand shocks to
the monetary system. From the standpoint of the monetary authority, demand
shocks are much more apt to be unanticipated under the current system than
are supply shocks (obviously). As such, unanticipated demand shocks are apt
to be small and are handled by automatic and very short term adjustments of
borrowed reserves, while supply shocks (or changes since folks at the Fed
these days don't really like it when their actions are described as shocks)
come from changing non-borrowed reserves. A simple inspection of the ratio
of non-borrowed to borrowed reserves (absent Continental Illinois type
effects) makes the relative importance of these two reserve types clear for
their roles in monetary policy.

>Instead Mr Witte attempts to argue
>his case, as he often does, by the use of ridicule and diversion.
>In this instance he selects comments of secondary importance in
>the letter to focus the readers attention away from a position he
>cannot defend.

Oh dear, first I am accused of not providing specific references in
a post where I have clearly included all the important citations and more,
now I am accused of diversion away from points that I cannot defend,
although I have repeatedly done so. The original post was a polite letter
from a guy at the Fed trying to patiently explain what the important issue
really is. I feel his pain.

Mason Clark

unread,
Mar 22, 2002, 2:42:15 PM3/22/02
to
On 22 Mar 2002 16:53:21 GMT, mwi...@merle.acns.nwu.edu (Mark Patrick Witte) wrote:

First Mark typed:

> Astonishing revelations! Textbooks simplify, there is not perfect
>short run control of the monetary aggregates (or FFR either), the Fed can
>control the money the money supply over longer periods. Wow, this guy was a
>real whistle-blower!
>

then he typed:

> The original post was a polite letter
>from a guy at the Fed trying to patiently explain what the important issue
>really is. I feel his pain.

Which is it Mark?

(I suspect you were thinking Warren Mosler as the "whistle-blower."
The insider at the Fed is not likely to be called such.)

Mason C (why are "economists" defensive?)

William F Hummel

unread,
Mar 22, 2002, 4:50:56 PM3/22/02
to
On 22 Mar 2002 16:53:21 GMT, mwi...@merle.acns.nwu.edu (Mark
Patrick Witte) wrote:

>In article <qihm9ugbt6ra4g1eh...@4ax.com>,
>William F Hummel <wfhu...@attbi.com> wrote:

>>On 22 Mar 2002 03:39:30 GMT, mwi...@merle.acns.nwu.edu (Mark
>>Patrick Witte) wrote:
>>
>>> Astonishing revelations! Textbooks simplify, there is not perfect
>>>short run control of the monetary aggregates (or FFR either), the Fed can
>>>control the money the money supply over longer periods. Wow, this guy was a
>>>real whistle-blower!
>>>
>>Mr. Witte apparently isn't even embarrassed by ignoring the key
>>statement made in the letter, namely "in the short run the
>>causation runs more from money to reserves than vice versa",
>>which was the point at issue.
>
> Ah, I would like to thank Mr. Hummel for letting me know which of
>Donald Kohn's words he considered the key ones. Since I'm not him, and lack
>mindreading ability, I would not have known.

Your welcome. Some need more help than others.

>In fact, to me, it looked like
>a the sort of polite letter I often wrote in my days working in Washington
>to constituents who wrote in with empassioned but not entirely worked out
>thoughts about government policy. (I know such people who have this as part
>of their duties at the Fed as well.)

Don't flatter yourself. For all I know you may be a star in
academia, but I doubt that Greenspan has ever heard of you.
Donald Kohn has had a distinguished career at the Fed. And you
claim he would be assigned a flunky job of the type you did at
the Fed. Checking Google, I found the following release by the
Fed BOG on Donald Kohn:

Quote

Release Date: June 11, 2001

For immediate release

The Federal Reserve Board today announced the appointments of
Donald Kohn as Advisor to the Board for Monetary Policy in the
Office of Board Members, Vincent Reinhart as Director of the
Division of Monetary Affairs and Brian Madigan as a Deputy
Director of the Division of Monetary Affairs, all effective July
2, 2001.

Mr. Kohn will continue as secretary of the Federal Open Market
Committee (FOMC), with responsibility for briefing the Committee
and for its announcements, minutes and transcripts. He holds a
Ph.D. from the University of Michigan and served as financial
economist for the Federal Reserve Bank of Kansas City before
joining the Board's staff in 1975. Mr. Kohn has been the Director
of the Division of Monetary Affairs since 1987 but is
relinquishing management of the division and will focus on issues
related to monetary policy.

Unquote



>Without Mr. Hummel's insight, I would
>have thought the letter was doing the classic, "But of course what you say
>is right (to some degree), *however* the more important issue is that..."
>In this case, the "however" falls right after the line cited by Mr. Hummel:
>
>>>>However, the Federal Reserve can control the composition of
>>>>reserves between non-borrowed and borrowed reserves fairly
>>>>closely in the short run.

The key words are those related to the question you asked in
another post:

"Where is "the Fed" on record with such a statement?"

The statement was:

"On the other hand if you are referring to all transaction money,
most of which is in the form of bank deposits, then even the Fed
will tell you it cannot exogenously change that supply."

And the answer is in the words I pointed to in the letter. It
really shouldn't require so much back and forth to answer your
question and put the matter to rest.


>
> From a technical standpoint, there are supply and demand shocks to
>the monetary system. From the standpoint of the monetary authority, demand
>shocks are much more apt to be unanticipated under the current system than
>are supply shocks (obviously). As such, unanticipated demand shocks are apt
>to be small and are handled by automatic and very short term adjustments of
>borrowed reserves, while supply shocks (or changes since folks at the Fed
>these days don't really like it when their actions are described as shocks)
>come from changing non-borrowed reserves. A simple inspection of the ratio
>of non-borrowed to borrowed reserves (absent Continental Illinois type
>effects) makes the relative importance of these two reserve types clear for
>their roles in monetary policy.

Much of what Mr. Witte said in the above paragraph is correct.
Unfortunately he also demonstrated how misinformed an economist
can be about monetary affairs when that is not his field. The
ratio of non-borrowed to borrowed reserves says nothing about
their relative importance as Mr. Witte asserts. The Fed must
keep the banking system short of non-borrowed reserves in order
to force some banks to borrow at the discount window. That's
necessary for the Fed to maintain control of the overnight
lending rate. But it doesn't take much borrowing at the window
to make that work. The Fed does not want to force banks to
borrow any more than necessary at the window. It should be
remembered that, in contrast to most other central banks, the Fed
does not pay interest on the reserve deposits of banks, which
puts US banks at a competitive disadvantage vis-a-vis most
foreign banks. Paying interest to the Fed in order to meet
reserve requirements would just compound that disadvantage.

>
>>Instead Mr Witte attempts to argue
>>his case, as he often does, by the use of ridicule and diversion.
>>In this instance he selects comments of secondary importance in
>>the letter to focus the readers attention away from a position he
>>cannot defend.
>
> Oh dear, first I am accused of not providing specific references in
>a post where I have clearly included all the important citations and more,

But I didn't see any footnotes.

>now I am accused of diversion away from points that I cannot defend,
>although I have repeatedly done so.

If you could defend them, there would have been no point in the
diversion.

>The original post was a polite letter
>from a guy at the Fed trying to patiently explain what the important issue
>really is. I feel his pain.

Yeah, a "polite" letter written by some "guy" at the Fed named
Donald Kohn. Sheesh!

WFH

Mark Patrick Witte

unread,
Mar 22, 2002, 6:52:55 PM3/22/02
to
In article <uv1n9uglna5um4sa4...@4ax.com>,

Mason Clark <mas...@ix.netcom.com> wrote:
>On 22 Mar 2002 16:53:21 GMT, mwi...@merle.acns.nwu.edu (Mark Patrick Witte) wrote:
>
>First Mark typed:
>
>> Astonishing revelations! Textbooks simplify, there is not perfect
>>short run control of the monetary aggregates (or FFR either), the Fed can
>>control the money the money supply over longer periods. Wow, this guy was a
>>real whistle-blower!

I hoped that my over-the-top presentation would indicate to readers
a certain degree of sarcasm. I was refering to specific elements in the
extensive quote from Donald Kohn, an employee of the Federal Reserve, who
wrote a letter that basically confirms the conventional wisdom about
monetary theory.

>then he typed:
>
>> The original post was a polite letter
>>from a guy at the Fed trying to patiently explain what the important issue
>>really is. I feel his pain.

Again, from Donald Kohn, an employee of the Federal Reserve, who
wrote a letter that basically confirms the conventional wisdom about
monetary theory.

>Which is it Mark?

Both refer to the remarks of Donald Kohn, an employee of the Federal
Reserve, who wrote a letter that basically confirms the conventional wisdom
about monetary theory.

>(I suspect you were thinking Warren Mosler as the "whistle-blower."
>The insider at the Fed is not likely to be called such.)

I have met Warren Mosler, but here I was refering to the
words of Donald Kohn, and employee of the Federal Reserve, who wrote
a letter that basically confirms the conventional wisdom about monetary
theory.

Mark Patrick Witte

unread,
Mar 22, 2002, 8:18:10 PM3/22/02
to
In article <ed9n9u852nlbf2mo0...@4ax.com>,

We have met, and work from my research group at one point led him to
change his testimony before Congress.

>Donald Kohn has had a distinguished career at the Fed. And you
>claim he would be assigned a flunky job of the type you did at
>the Fed.

Ah, the unwarrented assumptions people make when they do not read
closely. My work at the Federal Reserve did not involve writing such
letters, and those who did were senior economists, ironically. Further, I
never worked for the Board, and I doubt that anyone at the Fed views the
letter writing public as "constituents."

(I should also add that the letters I wrote when I was in DC
often went out over a signature other than my own, often that of a
senior staff person; not that the person who received the letter would
know this, nor would anyone else who latter happened to see, and perhaps
post such a letter to some newgroup. On the other hand, many senior staff
people also wrote such letters, so it's just hard to say. Either way,
the letters were like this one, and generally expressed sympathy with the
other person's point of view and then outlined the conventional wisdom.
An exception to this comes to mind when a friend of mine wrote a line in
a letter that got through the vetting process, but then proved quite
quotable by someone who read it in a very well televised event.)

>Checking Google, I found the following release by the
>Fed BOG on Donald Kohn:
>
>Quote
>
>Release Date: June 11, 2001
>
>For immediate release
>
>The Federal Reserve Board today announced the appointments of
>Donald Kohn as Advisor to the Board for Monetary Policy in the
>Office of Board Members, Vincent Reinhart as Director of the
>Division of Monetary Affairs and Brian Madigan as a Deputy
>Director of the Division of Monetary Affairs, all effective July
>2, 2001.
>
>Mr. Kohn will continue as secretary of the Federal Open Market
>Committee (FOMC), with responsibility for briefing the Committee
>and for its announcements, minutes and transcripts. He holds a
>Ph.D. from the University of Michigan and served as financial
>economist for the Federal Reserve Bank of Kansas City before
>joining the Board's staff in 1975. Mr. Kohn has been the Director
>of the Division of Monetary Affairs since 1987 but is
>relinquishing management of the division and will focus on issues
>related to monetary policy.
>
>Unquote

I'm sure that Dr. Kohn is a fine economist and a valuable player in
US monetary policy. Based upon the single piece of his writing that I have
ever read, I certainly have no reason to doubt his understanding of the
monetary system (if only I could say this of the Chairman). However, I am
confused, does Mr. Hummel feel that Kohn is a closet radical monetary
theorist? Or that he holds to some true, secret wisdom, that those inside
the Board know about but don't share with their academic colleagues (and in
fact have brain-washed out of them when they leave the BOG)? (I am not
saying that the Fed does not have this power, I am just asking if Mr. Hummel
believes they use it in this fashion. And yes, it does involve a small
device like the one from Men in Black.)



>>Without Mr. Hummel's insight, I would
>>have thought the letter was doing the classic, "But of course what you say
>>is right (to some degree), *however* the more important issue is that..."
>>In this case, the "however" falls right after the line cited by Mr. Hummel:
>>
>>>>>However, the Federal Reserve can control the composition of
>>>>>reserves between non-borrowed and borrowed reserves fairly
>>>>>closely in the short run.
>
>The key words are those related to the question you asked in
>another post:
>
>"Where is "the Fed" on record with such a statement?"
>
>The statement was:
>
>"On the other hand if you are referring to all transaction money,
>most of which is in the form of bank deposits, then even the Fed
>will tell you it cannot exogenously change that supply."
>
>And the answer is in the words I pointed to in the letter. It
>really shouldn't require so much back and forth to answer your
>question and put the matter to rest.

In the words of famed central banker Donald Kohn:

"However, the Federal Reserve can control the composition of
reserves between non-borrowed and borrowed reserves fairly

closely in the short run." (And by implication, in the long run as well.)

Ah yes, what do I know about monetary affairs? It's
not as though I've ever studied the subject, known a central
banker, worked as a consultant for a private financial firm,
or been on the payroll of the Federal Reserve System. However,
I feel it is horridly credentialist for someone to point this out.
I would gladly defer to someone who truely is an expert in
monetary affairs, perhaps an Alan Blinder, or a Larry Christiano,
or...maybe Donald Kohn, who points out that the Fed is able to
control bank reserves in the short run.

Anyway this is a simple economic question. It is not
clear what it means to "keep the system" short of reserves since
there is a scarcity value to reserves that creates the interest
rate on lending them, thus it's not clear why the Fed must worry
about "maintaining control of the overnight lending rate." (If
it went out of control...what would happen? What does it even
mean for a rate to be out of control? If they really want it
to be under control, the Fed could make it like the discount
rate, a take it or leave it rate for interbank borrowing or lending.)

I really don't want to be unfair to Mr. Hummel. I understand
that he has spent some years now reading and thinking a lot about
these issues and is a smart guy, but I have a hard time interpreting
some of his remarks. Maybe it's him, quite honestly, maybe it's me.

Mr. Hummel is right that under the Fed's current operating
procedure, it prefers to supply reserves trough open market operations,
and at least partially for the reason he explain. (However, the Fed's main
concern with the discount window is that banks would want to borrow too
much, give that the discount rate is almost always below the FFR.)

So, while what Mr. Hummel says is basically true, it is also
irrelevant to how the Fed pursues its monetary policy goals. I'm afraid
that here Mr. Hummel is confusing tactics with strategy. As such, the Fed
trys to affect the economy by managing the Fed Funds Rate, which implies
that it has to control the stock of reserves since this stock and the FFR
are co-determined.

>>>Instead Mr Witte attempts to argue
>>>his case, as he often does, by the use of ridicule and diversion.
>>>In this instance he selects comments of secondary importance in
>>>the letter to focus the readers attention away from a position he
>>>cannot defend.
>>
>> Oh dear, first I am accused of not providing specific references in
>>a post where I have clearly included all the important citations and more,
>
>But I didn't see any footnotes.

Ah! Damn UNIX editors, dropping all my ibids and op cits!
Well, one more time then.

[Larry Christiano about Japan
http://minneapolisfed.org/research/qr/qr1322.html
"Stochastic Trends and Economic Fluctuations" by Robert G. King,
Charles I. Plosser, James H. Stock, Mark W. Watson,
from the The American Economic Review, Vol. 81, No. 4. (Sep., 1991)
"Trends and Random Walks in Macroeconmic (sic) Time Series: Some
Evidence and Implications" by Nelson and Plosser, Journal of
Monetary Economics, 1982.
Solow's 1956 QJE paper or his 1957 RE Stat paper
Charles I. Jones, _Introduction to Economic Growth_."
(Norton Press, 2nd edition, 2002).
Stock and Watson, The Journal of Economic Perspectives, Vol. 2, No.
3. (Summer, 1988), pp. 147-174
Christiano and Eichenbaum, 1989 NBER Working Paper w3130
Thomas Kuhn, 1962, _The Structure of Scientific Revolutions_
_Central Banking in Theory and Practice_, Alan S. Blinder,
MIT Press, 1999
Cagan, Journal of Money Credit and Banking, 1969]

(Hmm, that might make a good reading list for the next time I teach money
and banking.)

>>now I am accused of diversion away from points that I cannot defend,
>>although I have repeatedly done so.
>
>If you could defend them, there would have been no point in the
>diversion.

Yeah, but when someone pushes me on this whole "co-determined"
thing, I clearly just run off and start new threads.

>>The original post was a polite letter
>>from a guy at the Fed trying to patiently explain what the important issue
>>really is. I feel his pain.
>
>Yeah, a "polite" letter written by some "guy" at the Fed named
>Donald Kohn. Sheesh!

I hope that Mr. Hummel is not implying that Donald Kohn is ever not
polite, or that...heaven forbid!, that he is not a guy. I mean, I know
someone in the world of economics who was named Donald but who is not
currently a guy, but I did not know there was becoming a trend, not that
there's anything wrong with it, no, no.

>WFH

William F Hummel

unread,
Mar 22, 2002, 11:55:15 PM3/22/02
to
On 23 Mar 2002 01:18:10 GMT, mwi...@merle.acns.nwu.edu (Mark
Patrick Witte) wrote:

(some stuff deleted)

>In article <ed9n9u852nlbf2mo0...@4ax.com>,
>William F Hummel <wfhu...@attbi.com> wrote:

>>On 22 Mar 2002 16:53:21 GMT, mwi...@merle.acns.nwu.edu (Mark
>>Patrick Witte) wrote:
>>
> I'm sure that Dr. Kohn is a fine economist and a valuable player in
>US monetary policy. Based upon the single piece of his writing that I have
>ever read, I certainly have no reason to doubt his understanding of the
>monetary system (if only I could say this of the Chairman). However, I am
>confused, does Mr. Hummel feel that Kohn is a closet radical monetary
>theorist? Or that he holds to some true, secret wisdom, that those inside
>the Board know about but don't share with their academic colleagues (and in
>fact have brain-washed out of them when they leave the BOG)? (I am not
>saying that the Fed does not have this power, I am just asking if Mr. Hummel
>believes they use it in this fashion. And yes, it does involve a small
>device like the one from Men in Black.)

Since I can't make any sense of this rambling discourse, I'd best
just pass the chance.


>
>>The key words are those related to the question you asked in
>>another post:
>>
>>"Where is "the Fed" on record with such a statement?"
>>
>>The statement was:
>>
>>"On the other hand if you are referring to all transaction money,
>>most of which is in the form of bank deposits, then even the Fed
>>will tell you it cannot exogenously change that supply."
>>
>>And the answer is in the words I pointed to in the letter. It
>>really shouldn't require so much back and forth to answer your
>>question and put the matter to rest.
>
> In the words of famed central banker Donald Kohn:
>
>"However, the Federal Reserve can control the composition of
>reserves between non-borrowed and borrowed reserves fairly
>closely in the short run." (And by implication, in the long run as well.)
>

Here Mr. Witte shows his confusion about the matter under
discussion. He has switched the focus from controlling the total
of reserves to controlling the composition of reserves. I
shouldn't have to remind him that these are two entirely
different types of challenges. The Fed can fairly easily control
the composition between borrowed and non-borrowed reserves, but
has it no control over the total of reserves for the reasons I've
outlined in some detail.

>>>
>>Much of what Mr. Witte said in the above paragraph is correct.
>>Unfortunately he also demonstrated how misinformed an economist
>>can be about monetary affairs when that is not his field. The
>>ratio of non-borrowed to borrowed reserves says nothing about
>>their relative importance as Mr. Witte asserts. The Fed must
>>keep the banking system short of non-borrowed reserves in order
>>to force some banks to borrow at the discount window. That's
>>necessary for the Fed to maintain control of the overnight
>>lending rate. But it doesn't take much borrowing at the window
>>to make that work. The Fed does not want to force banks to
>>borrow any more than necessary at the window. It should be
>>remembered that, in contrast to most other central banks, the Fed
>>does not pay interest on the reserve deposits of banks, which
>>puts US banks at a competitive disadvantage vis-a-vis most
>>foreign banks. Paying interest to the Fed in order to meet
>>reserve requirements would just compound that disadvantage.
>
> Ah yes, what do I know about monetary affairs? It's
>not as though I've ever studied the subject, known a central
>banker, worked as a consultant for a private financial firm,
>or been on the payroll of the Federal Reserve System.

None of which guarantees that you fully understand the system.

>However,
>I feel it is horridly credentialist for someone to point this out.
>I would gladly defer to someone who truely is an expert in
>monetary affairs, perhaps an Alan Blinder, or a Larry Christiano,
>

Don't feel bad. But credentials, even for very bright people,
are no guarantee of expertise. I don't know about Christiano,
but I am convinced that Blinder has his blind spots in regard to
monetary affairs. Surprisingly one doesn't necessarily develop a
deep understanding of the monetary system by working at the Fed.

>or...maybe Donald Kohn, who points out that the Fed is able to
>control bank reserves in the short run.

Not true at all. Kohn did not point out that the Fed is able to
control bank reserves in the short run. Here is what he said:

"Indeed, as you stress, in the short run the causation runs more


from money (see note below) to reserves than vice versa.

However, the Federal Reserve can control the composition of
reserves between non-borrowed and borrowed reserves fairly

closely in the short run. By varying this mix the Federal


Reserve can influence the federal funds rate and the other money
market rates in the short run and thereby affect growth of the
money supply over a longer period."

Note that he was talking about the composition of reserves, not
the total quantity of reserves, a distinction Mr. Witte failed to
grasp. Further, he used the qualifying term "fairly closely"
which Mr. Witte omitted. Two errors on one play will usually get
a guy replaced on the team. As noted above, the Fed can control
the composition. It cannot control the total.


>
> Anyway this is a simple economic question. It is not
>clear what it means to "keep the system" short of reserves

Again more reading confusion on Mr. Witte's part. The Fed does
not keep the system short of reserves. Quite the contrary. It
must supply whatever reserves banks need to meet their reserve
ratio requirement on checkable deposits.

The Fed does keep the system short of _unborrowed_ reserves for
the reason I mentioned earlier, namely to force some banks to the
discount window. This is a necessary part of controlling the
average interbank lending rate.

>since
>there is a scarcity value to reserves that creates the interest
>rate on lending them, thus it's not clear why the Fed must worry
>about "maintaining control of the overnight lending rate." (If
>it went out of control...what would happen? What does it even
>mean for a rate to be out of control?

The Fed funds rate was out of control during the early 1980s.
The resulting high volatility together with the high average Fed
funds rate was very difficult for the business community to deal
with. Neither the Fed nor the business community wants a repeat
of that.

>If they really want it
>to be under control, the Fed could make it like the discount
>rate, a take it or leave it rate for interbank borrowing or lending.)

You better believe they want the Fed funds rate under control.
And they use some sophisticated algorithms to assist in their OMO
to keep it there. It doesn't have to be smack on target, but it
is important to the whole financial world to have confidence that
the Fed will keep the overnight rate on the target that it has
announced. If it strays too far, it can suggest an unannounced
change in the funds rate. In 1992 the Fed pulled a surprise
increase of just .25% that stunned the bond market.

>
> I really don't want to be unfair to Mr. Hummel. I understand
>that he has spent some years now reading and thinking a lot about
>these issues and is a smart guy, but I have a hard time interpreting
>some of his remarks. Maybe it's him, quite honestly, maybe it's me.
>
> Mr. Hummel is right that under the Fed's current operating
>procedure, it prefers to supply reserves trough open market operations,
>and at least partially for the reason he explain. (However, the Fed's main
>concern with the discount window is that banks would want to borrow too
>much, give that the discount rate is almost always below the FFR.)

The Fed does not allow banks to borrow from the discount window
in order to lend at a profit. Further, the Fed has made it clear
to banks that borrowing from the discount window is not looked
upon favorably because it implies poor management. For this
reason, even though the discount rate is typically about 50 basis
point cheaper than the Fed funds target rate, banks themselves
are reluctant to borrow from the Fed unless the interbank lending
rate is simply too steep for them. This means there is a sloping
demand curve for Fed loans which the Fed has calibrated and can
use to control the composition of reserves (borrowed vs
unborrowed) which will hold the average interbank lending rate
close to target.

> So, while what Mr. Hummel says is basically true, it is also
>irrelevant to how the Fed pursues its monetary policy goals. I'm afraid
>that here Mr. Hummel is confusing tactics with strategy. As such, the Fed
>trys to affect the economy by managing the Fed Funds Rate, which implies
>that it has to control the stock of reserves since this stock and the FFR
>are co-determined.

There's that buzzword "co-determined" again, saying nothing but
sounding profound. Mr. Hummel is well aware of the distinction
between the art of controlling the Fed funds rate and the reason
for doing so at some particular rate. Surprisingly Mr. Witte
continues to misunderstand some of the key aspects of the
monetary system. The Fed has only one monetary policy tool, and
that is control of the Fed funds rate. Unless Mr. Witte can
overcome the notion that the Fed has to control (or can control)
the stock of banking system reserves it will be an albatross
around his neck. It will prevent him or anyone else interested
from an in depth understanding how and why the monetary system
works the way it does.

>
>>>The original post was a polite letter
>>>from a guy at the Fed trying to patiently explain what the important issue
>>>really is. I feel his pain.
>>
>>Yeah, a "polite" letter written by some "guy" at the Fed named
>>Donald Kohn. Sheesh!
>
> I hope that Mr. Hummel is not implying that Donald Kohn is ever not
>polite, or that...heaven forbid!, that he is not a guy. I mean, I know
>someone in the world of economics who was named Donald but who is not
>currently a guy, but I did not know there was becoming a trend, not that
>there's anything wrong with it, no, no.

I have little doubt that Donald Kohn understands the system in
depth and that he was pulling his punches in his response to
Mosler. Why? Because it avoids putting him on the spot. It
would upset the vast majority of economists, including many at
the Fed, who just can't swallow the notion that Fed cannot
control either the money supply or the reserves backing that
supply, except quite indirectly and with a long time lag.

WFH

Robert Vienneau

unread,
Mar 23, 2002, 5:17:07 AM3/23/02
to
William Hummel wrote:

"I would be interested to see an example of an economic trend
based purely on theory that has been validated empirically."

Mark Witte replied in some vague way:

"Well, take a look at Solow's work..."

Later he clarified he meant:

"Solow's 1956 QJE paper or his 1957 RE Stat paper..."

The complete reference for the latter paper is:

R. M. Solow, "Technical Change and the Aggregate Production Function".
Review of Economics and Statistics. August 1957.

Now Solow has written:

"The factor-share device of my 1957 article is in no sense a TEST
of aggregate production functions or marginal productivity or of
anything else. It merely shows how one goes about interpreting
given time series if one starts by ASSUMING that they were
generated from a production function and that the competitive
marginal-product relations apply."
-- R. M. Solow, "Laws of Production and Laws of Algebra: The
Humbug Production Function: A Comment". V. 56, 1. Feb. 1974,
p. 121.

Somewhat confusingly, Solow (1957) contains correlation coefficients.
So it seems Solow (1957) does not empirically validate anything.
I don't know if the other references Mr. Witte gave address Franklin
Fisher's simulation results or contrast the empirical strength of
Solow's theory with competing theories (e.g., Kaldor's).

Does Solow (1957) provide an economic trend based purely on theory?
Well, Samuelson has described this work as like that of a "busman
on holiday who operates brilliantly and without inhibitions in the
rough-and-ready realm of empirical heuristics". (This is in a paper
trying to provide some theoretical rationalization for Solow's
results; Samuelson later dropped this paper's defense.) And Hahn
has written:

"...Suppose that in fact there is a number C (a measure of
capital) defined by

C = C(R, xbar )

[where R is the rate of interest and xbar is a vector of endowments;
notation simplified.] If now we treat as given C and xbar then we
shall have one equation too many, since the information is equivalent
to being told R...

Lastly, suppose that we are given C and nothing else. Then the
investment savings equation can be written as

Z(R, C) = 0

which now, since C is fixed, is one equation in one unknown. The
condition that the ...market [for endowments] also clears and the
definition

C = pbar( R ) . xbar

[where pbar( R ) is a vector of prices and . is the inner product]
then give us what the amount of [endowments]...would have to be in
order to have a consistent story...The problem is the sense to be
made of C being given from the outside.

...In general, there does not exist a function from the vector of
endowments to the scalar such that knowledge of the scalar (and of
preferences and of technology) is sufficient to allow one to
determine a neoclassical equilibrium. If you put it the other way
round, it is even more obvious. In general, the neoclassical
equilibrium can be found given the vector of endowments which may
have, say, 10^8 components. It would be surprising if there were
a single number which gives the same information as the 10^8
dimensional vector. In fact, sometimes and in very special cases,
this surprising property holds. But neoclassical economists have
shown these special cases to be without interest."
-- Frank Hahn

So Mr. Witte seems to be citing a description of an economic trend that
is not based purely on theory and that does not seem to have been
empirically validated. At least, that is how the pure-reviewed
literature, including more recent references (e.g., by Burmeister)
characterizes the work in his citation. I don't know why Mr. Witte
disagrees. Mayhaps he read William as asking for a heuristic,
empirically unvalidated and theoretically unlikely, useful in
interpretingand summarizing empirical data.

--
Try http://csf.colorado.edu/pkt/pktauthors/Vienneau.Robert/Bukharin.html
To solve Linear Programs: .../LPSolver.html
r c A game: .../Keynes.html
v s a Whether strength of body or of mind, or wisdom, or
i m p virtue, are found in proportion to the power or wealth
e a e of a man is a question fit perhaps to be discussed by
n e . slaves in the hearing of their masters, but highly
@ r c m unbecoming to reasonable and free men in search of
d o the truth. -- Rousseau

George

unread,
Mar 23, 2002, 2:28:33 PM3/23/02
to
wfhummel wrote in part:

>The Fed can fairly easily control
the composition between borrowed and non-borrowed reserves, but
has it no control over the total of reserves for the reasons I've
outlined in some detail.>

Could one of you guys when you have nothing else to do, explain "borrowed
reserves"? Banks keep reserves so they do not not have a run they can't handle.
If banks borrow reserves isn't this kinda circumventing the purpose of
reserves. If a bank doesn't have enough reserves, that's bad. Just curious.

William F Hummel

unread,
Mar 23, 2002, 3:50:17 PM3/23/02
to

Borrowed reserves are those funds obtained on loan from the Fed,
which of course must be repaid at some point. The Fed charges
interest (at the so-called discount rate) on those loans.

Non-borrowed reserves are those funds that flow to banks as
customers deposit the proceeds they receive from the Fed when it
purchases securities (mainly Treasury securities) from the
public. This comprises 90% to 95% of total reserves.

Banks borrow all the time and from many sources, not simply from
the Fed. Large banks borrow enormous amounts in the money market
to support their own lending. However only those funds obtained
from the Fed (borrowed or unborrowed) can increase the aggregate
reserves of the banking system. All other borrowing just shifts
ownership of reserves from one bank to another. So when banks in
the aggregate are faced with a shortage of reserves to meet their
reserve ratio requirements, one or more banks will have to borrow
from the Fed. And the Fed has no real choice but to grant the
loan.

For more detail, see http://wfhummel.cnchost.com/moneybasics.html
and http://wfhummel.cnchost.com/bankreserves.html

WFH

Grinch

unread,
Mar 23, 2002, 4:17:20 PM3/23/02
to
On 23 Mar 2002 19:28:33 GMT, gchan...@aol.com (George) wrote:

>wfhummel wrote in part:
>
>>The Fed can fairly easily control
>the composition between borrowed and non-borrowed reserves, but
>has it no control over the total of reserves for the reasons I've
>outlined in some detail.>
>
>Could one of you guys when you have nothing else to do, explain "borrowed
>reserves"? Banks keep reserves so they do not not have a run they can't handle.

No, deposit insurance exists to prevent/handle bank runs.

Reserve requirements are a tool of monetary policy.

E.g., see http://www.ny.frb.org/pihome/fedpoint/fed45.html

>If banks borrow reserves isn't this kinda circumventing the purpose of
>reserves. If a bank doesn't have enough reserves, that's bad. Just curious.

Since that's not the purpose of the reserve requirement, it's not a
problem.

Remember, banks must keep required reserves on hand or on deposit at
the Fed -- so they aren't available to pay depositors in event of a
bank run. After all, they can't be on hand or on deposit at the Fed
and be being paid out at the same time.

Simplistic example: A bank takes $100 of deposits and after applying
the current 10% reserve requirement loans out $90, leaving it with $10
of reserves on hand. Now there is a "mini bank run" during which
depositors withdraw $5 -- only 5% of deposits but 50% of the reserves.
That leaves the bank with too little in reserves. But it can borrow
some to make up the shortfall (and avoid having to call in loans to do
so instead) until it attracts some more deposits or reduces its loans
outstanding as they mature without calling any in, as explained in the
various "Fed Points".


Grinch

unread,
Mar 23, 2002, 5:14:36 PM3/23/02
to
On Sat, 23 Mar 2002 20:50:17 GMT, William F Hummel
<wfhu...@attbi.com> wrote:

>On 23 Mar 2002 19:28:33 GMT, gchan...@aol.com (George) wrote:
>
>>wfhummel wrote in part:
>>
>>>The Fed can fairly easily control
>>the composition between borrowed and non-borrowed reserves, but
>>has it no control over the total of reserves for the reasons I've
>>outlined in some detail.>
>>
>>Could one of you guys when you have nothing else to do, explain "borrowed
>>reserves"? Banks keep reserves so they do not not have a run they can't handle.
>>If banks borrow reserves isn't this kinda circumventing the purpose of
>>reserves. If a bank doesn't have enough reserves, that's bad. Just curious.
>
>Borrowed reserves are those funds obtained on loan from the Fed,
>which of course must be repaid at some point. The Fed charges
>interest (at the so-called discount rate) on those loans.

Borrowed reserves more typically are borrowed from other banks,
financial institutions, government agencies, securities dealers, etc.,
through the federal funds market, with fed funds rate being the price
of such borrowing.

Reserves borrowed in the fed funds market can be used to support
for-profit lending, while those borrowed from the discount window
can't be -- one very practical reason why banks prefer to borrow
reserves in the fed funds market, even though the discount window
usually charges a lower interest rate, so borrowing from the discount
window usually remains slight.

>Non-borrowed reserves are those funds that flow to banks as
>customers deposit the proceeds they receive from the Fed when it
>purchases securities (mainly Treasury securities) from the
>public. This comprises 90% to 95% of total reserves.
>
>Banks borrow all the time and from many sources, not simply from
>the Fed. Large banks borrow enormous amounts in the money market
>to support their own lending. However only those funds obtained
>from the Fed (borrowed or unborrowed) can increase the aggregate
>reserves of the banking system. All other borrowing just shifts
>ownership of reserves from one bank to another. So when banks in
>the aggregate are faced with a shortage of reserves to meet their
>reserve ratio requirements, one or more banks will have to borrow
>from the Fed. And the Fed has no real choice but to grant the
>loan.

It seems the Fed must have the real choice to say "no" to loan
requests at the discount window, and must usually be expected to.
Otherwise, being that the discount rate is usually lower than the fed
funds rate, banks that are short of reserves would be lined up to
borrow at the discount window instead of in the fed funds market.

Even if "aggregate" reserves were deemed too low, this would be
reflected in the fed funds rate being "too high" and the typical
response by the Fed would be to choose to increase aggregate reserves
through open market operations rather than make loans upon request at
the discount window.

The Fed generally sets rather restrictive conditions for loans made
through the discount window, which is why there are so few of them.
But it has on particular occasions, such as after 9/11 and for Y2k,
announced the easing of these conditions to make it easier for banks
to borrow at the window for limited periods of time -- something which
wouldn't make much sense if it had no real choice but to grant the
loan requests to begin with.

See
http://www.ny.frb.org/pihome/fedpoint/fed15.html and
http://www.ny.frb.org/pihome/fedpoint/fed18.html

William F Hummel

unread,
Mar 23, 2002, 7:12:10 PM3/23/02
to
On Sat, 23 Mar 2002 17:14:36 -0500, Grinch
<oldn...@mindspring.com> wrote:

>On Sat, 23 Mar 2002 20:50:17 GMT, William F Hummel
><wfhu...@attbi.com> wrote:
>
>>On 23 Mar 2002 19:28:33 GMT, gchan...@aol.com (George) wrote:
>>
>>>wfhummel wrote in part:
>>>
>>>>The Fed can fairly easily control
>>>the composition between borrowed and non-borrowed reserves, but
>>>has it no control over the total of reserves for the reasons I've
>>>outlined in some detail.>
>>>
>>>Could one of you guys when you have nothing else to do, explain "borrowed
>>>reserves"? Banks keep reserves so they do not not have a run they can't handle.
>>>If banks borrow reserves isn't this kinda circumventing the purpose of
>>>reserves. If a bank doesn't have enough reserves, that's bad. Just curious.
>>
>>Borrowed reserves are those funds obtained on loan from the Fed,
>>which of course must be repaid at some point. The Fed charges
>>interest (at the so-called discount rate) on those loans.
>
>Borrowed reserves more typically are borrowed from other banks,
>financial institutions, government agencies, securities dealers, etc.,
>through the federal funds market, with fed funds rate being the price
>of such borrowing.

Not true. The term "borrowed reserves" refers specifically to
funds borrowed from the Fed, not from other financial
institutions. The reason for this distinction is that borrowing
from the Fed increases the aggregate reserves, while borrowing in
the Fed funds market does not. Borrowing in the Fed funds market
is merely a temporary transfer between banks of the ownership of
existing reserves on deposit at the Fed. They are part of the
non-borrowed component of reserves.

>
>Reserves borrowed in the fed funds market can be used to support
>for-profit lending, while those borrowed from the discount window
>can't be -- one very practical reason why banks prefer to borrow
>reserves in the fed funds market, even though the discount window
>usually charges a lower interest rate, so borrowing from the discount
>window usually remains slight.

The Fed does not allow banks to borrow from the discount window
in order to lend at a profit, but it normally relies on the bank
to honor that rule. The Fed has made it clear to banks that


borrowing from the discount window is not looked upon favorably

because it suggests poor management. It's referred to as the
"frown cost" of borrowing from the Fed. For that reason, even


though the discount rate is typically about 50 basis point

cheaper than the Fed funds target rate, banks are reluctant to


borrow from the Fed unless the interbank lending rate is simply
too steep for them. This means there is a sloping demand curve
for Fed loans which the Fed has calibrated and can use to control
the composition of reserves (borrowed vs unborrowed) which will

hold the average interbank lending rate close to target. It's a
strange system with none other like it, but it works.

>
>>Non-borrowed reserves are those funds that flow to banks as
>>customers deposit the proceeds they receive from the Fed when it
>>purchases securities (mainly Treasury securities) from the
>>public. This comprises 90% to 95% of total reserves.
>>
>>Banks borrow all the time and from many sources, not simply from
>>the Fed. Large banks borrow enormous amounts in the money market
>>to support their own lending. However only those funds obtained
>>from the Fed (borrowed or unborrowed) can increase the aggregate
>>reserves of the banking system. All other borrowing just shifts
>>ownership of reserves from one bank to another. So when banks in
>>the aggregate are faced with a shortage of reserves to meet their
>>reserve ratio requirements, one or more banks will have to borrow
>>from the Fed. And the Fed has no real choice but to grant the
>>loan.
>
>It seems the Fed must have the real choice to say "no" to loan
>requests at the discount window, and must usually be expected to.
>Otherwise, being that the discount rate is usually lower than the fed
>funds rate, banks that are short of reserves would be lined up to
>borrow at the discount window instead of in the fed funds market.

The Fed expects banks to exhaust other borrowing options before
coming to the discount window for a loan. A bank that comes too
often will be "frowned upon" and likely denied access to the
window. That could force the bank to sell securities or pay a
much higher rate in the money market to acquire the needed
reserves. That's the main reason why banks don't line up at the
window. However if they have no other source for meeting their
reserve requirements they will have to request a loan from the
Fed. The Fed will almost always grant the loan, but at a penalty
above the Fed funds rate.


>
>Even if "aggregate" reserves were deemed too low, this would be
>reflected in the fed funds rate being "too high" and the typical
>response by the Fed would be to choose to increase aggregate reserves
>through open market operations rather than make loans upon request at
>the discount window.

True. Except that banks in need of reserves usually do not come
to the window until near the end of the two weeks maintenance
period on which their average reserves are counted. That can
leave inadequate time for the funds market to redistribute
reserves. Under those conditions the Fed will normally grant the
loan, and possibly at a penalty rate.

>
>The Fed generally sets rather restrictive conditions for loans made
>through the discount window, which is why there are so few of them.
>But it has on particular occasions, such as after 9/11 and for Y2k,
>announced the easing of these conditions to make it easier for banks
>to borrow at the window for limited periods of time -- something which
>wouldn't make much sense if it had no real choice but to grant the
>loan requests to begin with.

Except for seasonal demand in some banks like those serving the
farming business, the Fed allows healthy banks to borrow only for
the purpose of covering a short term reserve deficiency that
could not be covered by borrowing elsewhere.

WFH

Grinch

unread,
Mar 23, 2002, 8:55:28 PM3/23/02
to
On Sun, 24 Mar 2002 00:12:10 GMT, William F Hummel
<wfhu...@attbi.com> wrote:

I was addressing George's question, which was about, to the best of my
understanding, the situation where "a bank" borrows reserves.

To wit:
~~
"Banks keep reserves so they do not have a run they can't handle. If


banks borrow reserves isn't this kinda circumventing the purpose of

reserves? If a bank doesn't have enough reserves, that's bad. Just
curious."
~~
Note: "If a bank..." The perspective discussed is that of individual
banks.

Banks borrow reserves from each other. Most reserves that are
borrowed are borrowed in this manner. The fed funds rate is the
interest rate on these borrowed reserves, and it is commonly referred
to as such. Without even doing a google search I could cite several
examples of fed funds being referred to by the term "borrowed
reserves" just from my files. [1]

Now, from the perspective of the banking system as a whole, you are
of course correct. Fed funds in the aggregate represent no borrowed
reserves since what is borrowed is also lent after being first
deposited.

So when discussing the banking system in aggregate, the terminology
"borrowed reserves" refers as you say to discount window borrowing,
and other reserves are "non-borrowed" even if in fact they are
borrowed from the owner through the fed funds market.

But that was not what I was talking about, since, to the best of my
admittedly sometimes faulty understanding, George was asking about the
borrowed reserves of a bank, not of the banking system in aggregate.

So it seems the Fed can say "no"at the discount window, and expects
banks that can anticipate that response not to ask.

Well, there had to be some reason for banks that need to borrow
reserves not to go there to get the lower interest rate.

>>>Non-borrowed reserves are those funds that flow to banks as
>>>customers deposit the proceeds they receive from the Fed when it
>>>purchases securities (mainly Treasury securities) from the
>>>public. This comprises 90% to 95% of total reserves.
>>>
>>>Banks borrow all the time and from many sources, not simply from
>>>the Fed. Large banks borrow enormous amounts in the money market
>>>to support their own lending. However only those funds obtained
>>>from the Fed (borrowed or unborrowed) can increase the aggregate
>>>reserves of the banking system. All other borrowing just shifts
>>>ownership of reserves from one bank to another. So when banks in
>>>the aggregate are faced with a shortage of reserves to meet their
>>>reserve ratio requirements, one or more banks will have to borrow
>>>from the Fed. And the Fed has no real choice but to grant the
>>>loan.
>>
>>It seems the Fed must have the real choice to say "no" to loan
>>requests at the discount window, and must usually be expected to.
>>Otherwise, being that the discount rate is usually lower than the fed
>>funds rate, banks that are short of reserves would be lined up to
>>borrow at the discount window instead of in the fed funds market.
>
>The Fed expects banks to exhaust other borrowing options before
>coming to the discount window for a loan. A bank that comes too
>often will be "frowned upon" and likely denied access to the
>window.

It will hear "No." That would seem to be proof that the Fed does in
fact have a real choice but to grant the loan.

>That could force the bank to sell securities or pay a
>much higher rate in the money market to acquire the needed
>reserves. That's the main reason why banks don't line up at the
>window.

A whole lot of anticipated "Nos".

Yes, the Fed will readily grant loans at the discount window to any
bank that meets all the conditions that it sets for obtaining such a
loan, as discussed in Fed Point 18, linked to previously.

>WFH

[1] E.g.:

"The interest rate that is paid on these borrowed reserves is called
the federal funds rate."
http://www.clev.frb.org/research/com2001/1001.htm
That's a Fed publication.

and...
"The interest rate charged on the borrowed reserves is called the
federal funds rate.
"Today, the federal funds market is much larger than before.
Commercial banks, particularly large ones, borrow these funds for any
shortage of their required reserves...."
http://www.biz.uiowa.edu/iem/assignments/fedpolicyandIEM1.doc

So it depends on what one is talking about: The system in aggregate,
for which fed funds aren't borrowed at all; or the individual banks
that borrow them from other institutions to make up a shortage of
required reserves, for which they certainly are "borrowed reserves."


Mark Patrick Witte

unread,
Mar 23, 2002, 11:26:34 PM3/23/02
to
Long time fans of sci.econ know to watch for certain things.
With Robert Vienneau, things always seem to go poorly when he tries to
get involved with issues concerning support in the data for models. And
also, it's usually the case that what he edits out is far more revealing
than what he leaves in. Under these rules, this post is a two-for!

In article <rvien-B1302F....@news.dreamscape.com>,


Robert Vienneau <rv...@see.sig.com> wrote:
>William Hummel wrote:
>
> "I would be interested to see an example of an economic trend
> based purely on theory that has been validated empirically."

The discussion was one that began over issues in macroeconomics
that were not directly related to money/financial matters, with long run
growth models being suggested as an example of such. Hmm...what would
be an example of a model that might relate to such a thing an would
predict a trend, and one where there is some emprical validation? Gosh, who
has done research like that? Maybe...Solow!

>Mark Witte replied in some vague way:
>
> "Well, take a look at Solow's work..."

Ah, observe a classic Vienneau edit! So crisp, but what does it
leave out? What exactly did I say in reply to these words from Mr. Hummel?

W| "Well, take a look at Solow's work...or the paper I recommended
W| by Christiano." A nice place to read about empirical tests of Solow
W| is Charles I. Jones's book _Introduction to Economic Growth_.

Thus we see a classic example of Mr. Vienneau's approach to honest
debate on sci.econ.

I put Jones's book as a serious recommendation for (post DeLong)
empirical support for the Solow approach. I agree that Solow's papers
themselves are not themselves the place to go for empirical support of this
methodology, and so favor Jones's exceptionally readable book.

As such, the following comments are on par for relevancy with Mr.
Vienneau's usual inputs on empirical matters of theoretical models finding
support in observed features of the real world.

Mark Patrick Witte

unread,
Mar 24, 2002, 12:26:56 AM3/24/02
to
Mr. Hummel seems to have borrowed the aggressive editor from someone
else who posts on this thread and has left a number of questions I posed to
him unanswered. I'll try to answer his while recalling some of my previous
points, and be brief to the limits of my small ability on that score.

* Does Mr. Hummel think that Dr. Donald Kohn is a fellow traveler
who is keeping his maverick views on monetary policy a secret from his peers
there, as indicated in the letter Mr. Hummel has posted on this group? Or
perhaps was Dr. Kohn saying something that is quite tame and in accord with
standard monetary theory? When he says that the Fed "can control the


composition of reserves between non-borrowed and borrowed reserves

fairly closely in the short run", it seems to me that he is saying that the
Fed closely controls this ratio because it controls both borrowed (small)
and non-borrowed (large, and tightly controlled via OMOs) reserves. We
don't know exactly what remark Kohn was responding to from Mr. Mosler, so
we're left with an Occam's Razor matter here, and the simplest reading is
that Kohn is saying that the Fed controls the level of borrowed reserves
well in the short run. Perhaps when Mr. Hummel gets around to citing the
textbooks that he finds so in error, I suppose starting with anything by
Alan Blinder, he can also put up some pages from published Fed documents
that support his claim, rather than a possibly misread passage from a
letter that was in response to something unclear. Or perhaps Mr. Hummel
would like to clarify matters with a more in-depth letter of his own to
Dr. Kohn where he gets the Fed's official view on whether interest rates
and the monetary base are co-determined.

* Does Mr. Hummel agree that the Federal reserve has made a
conscious policy *choice* to manage the Fed Funds Rate and it could instead
*choose* to manage reserves? I suspect that everyone on sci.econ (with one
possible exception) agrees that this is a good policy, but it is a policy
choice, not a law of nature.

* Does Mr. Hummel agree that if A + B = 10, and if we choose A to
be 7, then B must be 3? That is, having chosen A, we are left with a
Hobson's choice for B? And if we were instead to choose B first, then we
would have also at that moment limited our choice for A to a single number?
That is, that A and B are co-determined in the specified relationship?

And now, I shall do my part:

>Please answer a simple question. Let's assume banks issue loans
>this month well in excess of the aggregate reserves required to
>back them. (It can happen.) Assume also the FFR target has been
>selected previously and is required to be accurately met by the
>NYFRB. Tell me how your static model of X and Y are consistent
>with this imbalance, or what happens to restore equilibrium.

As taught in Introductory Macroeconomics, since the Fed has
chosen to target the FFR, it will adjust the amount of reserves in
the system, either through expansionary OMOs, or through increased discount
lending. However, if the Fed where to instead target reserves, then
it would be the FFR that would bear the brunt of the adjustment, since these
variables are co-determined. However, in the real world, in either case,
the Fed would allow both variables to adjust to meet the demand shock for
reserves; that is, the FFR displays volitility around its target range and
the stock of private sector bank reserves has an elasticity that is
certainly not zero since the amount of bank reserves is less than 10% of the
available monetary base.


In article <8e2o9u0dni6q8s6hm...@4ax.com>,

Mason Clark

unread,
Mar 24, 2002, 2:18:32 AM3/24/02
to

I would be interested to see an example of an economic trend
based purely on theory that has been validated empirically.

And especially if it will predict the value of GE stock.

Robert Vienneau

unread,
Mar 24, 2002, 7:13:54 AM3/24/02
to
Mark Patrick Witte is being silly:

> Long time fans of sci.econ know to watch for certain things.
> With Robert Vienneau, things always seem to go poorly when he tries to
> get involved with issues concerning support in the data for models.

Ad hominem.

> And
> also, it's usually the case that what he edits out is far more revealing
> than what he leaves in. Under these rules, this post is a two-for!

Ad hominem.


> In article <rvien-B1302F....@news.dreamscape.com>,
> Robert Vienneau <rv...@see.sig.com> wrote:
> >William Hummel wrote:

> > "I would be interested to see an example of an economic trend
> > based purely on theory that has been validated empirically."

> The discussion was one that began over issues in macroeconomics
> that were not directly related to money/financial matters, with long run
> growth models being suggested as an example of such. Hmm...what would
> be an example of a model that might relate to such a thing an would
> predict a trend, and one where there is some emprical validation? Gosh,
> who has done research like that? Maybe...Solow!

I assume my readers have a memory.



> >Mark Witte replied in some vague way:

> > "Well, take a look at Solow's work..."

> Ah, observe a classic Vienneau edit! So crisp, but what does it
> leave out? What exactly did I say in reply to these words from Mr.
> Hummel?

> W| "Well, take a look at Solow's work...or the paper I recommended
> W| by Christiano." A nice place to read about empirical tests of Solow
> W| is Charles I. Jones's book _Introduction to Economic Growth_.

I've never been able to obtain:

Sylos Labini, Paolo. 1995. "Why the Interpretation of the Cobb-Douglas
Production Function Must be Radically Changed". Structural Change and
Economic Dynamics, 6: 485-504.

Sylos Labini, Paolo. 1988. "The Great Debates on the Laws of Return
and the Value of Capital: When will Economists Finally Accept Their
Own Logic?". BNL Quarterly Review, September.


"The reality is that estimations of aggregate production functions
tend to yield poor results. For example, a standard finding for the
Cobb-Douglas with a linear time trend is a negative elasticity of
capital. See Sylos Labini (1995)..."

"At the empirical level, and contrary to widespread belief,
production functions, when estimated econometrically, tend to yield,
in general, poor results, a point made recently by Sylos-Labini
(1995) discussing estimations with the Cobb-Douglas function."
-- Jesus Felipe and Franklin M. Fisher, "Aggregation in Production
Functions: What Applied Economists Should Know". September 19,
2001. (Available somewhere on the 'net.)

I think this paper asserts that Mr. Witte is "simply and plainly wrong".

> Thus we see a classic example of Mr. Vienneau's approach to honest
> debate on sci.econ.

Ad hominem.


> I put Jones's book as a serious recommendation for (post DeLong)

> empirical support for the Solow approach. I AGREE that Solow's papers


> themselves are not themselves the place to go for empirical support of
> this
> methodology, and so favor Jones's exceptionally readable book.

(emphasis added.) Failed attempts to find a theoretical justification
for Solow's models are to be found elsewhere than the two papers
Mr. Witte cites.



> As such, the following comments are on par for relevancy with Mr.
> Vienneau's usual inputs on empirical matters of theoretical models
> finding
> support in observed features of the real world.

Ad hominem.

In the post Mr. Witte is pretending to comment on, I wrote:

"I don't know if the other references Mr. Witte gave address Franklin
Fisher's simulation results or contrast the empirical strength of
Solow's theory with competing theories (e.g., Kaldor's)."

So obviously, I acknowledged that Mr. Witte gave other references.
The ad hominems constituting almost the entirety of Mr. Witte's
comments seem to be based on a strawperson.

Does Mr. Witte address the point that Solow's model is more of
an heuristic, than "an economic trend based purely on theory"?
Does he explain why he disagrees with Samuelson's characterization
of this work as like that of a "busman on holiday who operates


brilliantly and without inhibitions in the rough-and-ready realm of

empirical heuristics"? No.

Does he clarify whether the literature he recommends addresses the
difficulties that past researchers, such as Franklin Fisher, have
found with some work that could be read as an attempt to validate
Solow's model? No.

Does he clarify whether the literature he recommends compares and
contrasts the ability of Solow's model to explain the data with
the ability of other approaches to explain the data? No.

I guess all the ad hominems Mr. Witte produces are to distract
from the fact he has nothing substantial to say, other than to
agree with my point that those particular classic Solow papers
do not contain empirical validation.

Given the silliness in Mr. Witte's post, it is extremely odd that
he should feel justified in making any comments about others'


"approach to honest debate on sci.econ".

--

William F Hummel

unread,
Mar 24, 2002, 1:04:11 PM3/24/02
to
On 24 Mar 2002 05:26:56 GMT, mwi...@merle.acns.nwu.edu (Mark
Patrick Witte) wrote:

> Mr. Hummel seems to have borrowed the aggressive editor from someone
>else who posts on this thread and has left a number of questions I posed to
>him unanswered. I'll try to answer his while recalling some of my previous
>points, and be brief to the limits of my small ability on that score.
>
> * Does Mr. Hummel think that Dr. Donald Kohn is a fellow traveler

etc. etc.

Mr. Witte does love to ramble on. I've already stated my views
on the letter from Kohn and have no intention of belaboring the
issue. Answering a series of mostly rhetorical questions adds
nothing to this discussion.


>
> * Does Mr. Hummel agree that the Federal reserve has made a
>conscious policy *choice* to manage the Fed Funds Rate and it could instead
>*choose* to manage reserves? I suspect that everyone on sci.econ (with one
>possible exception) agrees that this is a good policy, but it is a policy
>choice, not a law of nature.

If Mr. Witte understood some of my earlier remarks about the
Fed's actions in the early 1980s, he would not have asked this
question. At least I would hope that's true.


>
> * Does Mr. Hummel agree that if A + B = 10, and if we choose A to
>be 7, then B must be 3? That is, having chosen A, we are left with a
>Hobson's choice for B? And if we were instead to choose B first, then we
>would have also at that moment limited our choice for A to a single number?
>That is, that A and B are co-determined in the specified relationship?

Of course, but this trivial example bears no relation to the
question of whether the Fed can simultaneously control the Fed
funds rate and the total amount of reserves as Mr. Witte has
claimed, based on his reading of introductory macro textbooks.
In fact it should be apparent that the two monetary variables are
not dimensionally additive. I'll have more to say about how the
Fed manages to control the overnight funds rate at another time.


>
> And now, I shall do my part:
>
>>Please answer a simple question. Let's assume banks issue loans
>>this month well in excess of the aggregate reserves required to
>>back them. (It can happen.) Assume also the FFR target has been
>>selected previously and is required to be accurately met by the
>>NYFRB. Tell me how your static model of X and Y are consistent
>>with this imbalance, or what happens to restore equilibrium.
>
> As taught in Introductory Macroeconomics, since the Fed has
>chosen to target the FFR, it will adjust the amount of reserves in
>the system, either through expansionary OMOs, or through increased discount
>lending. However, if the Fed where to instead target reserves, then
>it would be the FFR that would bear the brunt of the adjustment, since these
>variables are co-determined. However, in the real world, in either case,
>the Fed would allow both variables to adjust to meet the demand shock for
>reserves; that is, the FFR displays volitility around its target range and
>the stock of private sector bank reserves has an elasticity that is
>certainly not zero since the amount of bank reserves is less than 10% of the
>available monetary base.

Mr. Witte says that according to his reading of introductory
macroeconomics, the Fed controls either the FFR or reserves and
lets the other one adjust. [I would note here that is an example
of a control variable and a dependent variable.] Then he
modifies this by saying in the real world both variables adjust
together to meet the demand shock for reserves, although he
doesn't acknowledge that one is the target variable and the other
is a "don't care" variable. Apparently what is most important to
him is that this must be seen as "co-determination".

Now it is true that the Fed uses its control of reserves to steer
the Fed funds rate to its target value, a procedure I have
described in some detail on sci.econ. But if that is the whole
point, it is hardly worth the time and effort to say it. That's
a case of not seeing the forest for the trees. Or to put it more
directly, one cannot acquire a real understanding of how our
monetary system works without seeing the bigger relationships.

WFH

susupply

unread,
Mar 24, 2002, 3:15:25 PM3/24/02
to

"Robert Vienneau" <rv...@see.sig.com>

The Conscience of sci.econ,

wrote in message news:rvien-993721....@news.dreamscape.com...

> Mark Patrick Witte is being silly:

BECAUSE:

> > Long time fans of sci.econ know to watch for certain things.
> > With Robert Vienneau, things always seem to go poorly when he tries to
> > get involved with issues concerning support in the data for models.
>
> Ad hominem.

Which may break the J.M. Ivler record for; Self-Unawareness in the 60 Meter
Dash.

Of course, Vienneau is incorrect that the above from Prof. Witte is an "ad
hominem". Nor is he correct in the following:

> > And
> > also, it's usually the case that what he edits out is far more revealing
> > than what he leaves in. Under these rules, this post is a two-for!
>
> Ad hominem.

Accurate descriptions of a poster's practices, when apposite, are not ad
hominems. The truth can legitimately be used, even if it stings.

[snip]

> I assume my readers have a memory.

Well, that doesn't explain Hummel's inability to remember from one paragraph
to the next what he's saying. In fact, most of your fans operate in a
similar manner.

But, I'm sure Prof. Witte remembers other times when you've said similar to
the following:

> I've never been able to obtain:
>
> Sylos Labini, Paolo. 1995. "Why the Interpretation of the Cobb-Douglas
> Production Function Must be Radically Changed". Structural Change and
> Economic Dynamics, 6: 485-504.

Which makes MPW's:

> > Thus we see a classic example of Mr. Vienneau's approach to honest
> > debate on sci.econ.
>
> Ad hominem.

Not the immediately above charge, but an accurate description of Mr.
Vienneau's approach.

But speaking of economists who DO rely almost exclusively on ad hominem, how
about those Two Daffy Aussies, Robert is so fond of?

Who said:

> "WHY THE PUBLIC NEEDS TO KNOW THAT ECONOMICS IS INTELLECTUALLY UNSOUND."

and:

> "Virtually every aspect of conventional economic theory is intellectually
> unsound; virtually every economic policy recommendation is just as likely
to
> do general harm as it is to lead to the general good. Far from holding the
> intellectual high ground, economics rests on foundations of quicksand. If
> economics were truly a science, then the dominant school of thought in
> economics would long ago have disappeared from view."

Then there was his sidekick, John Legge, just today chirping in with:

> Dr Kates is in the interesting position of being on two horses at once.
> When his members are the targets of some of the more bloody-minded
industry
> regulators and competition czars I am sure that Dr Kates will fully
support
> the debunking of such lords of their private dunheaps. When he is called
> upon to support deregulation of the labour market he would like to believe
> that, in spite of the odd lacunae, the neoclassical arguments for
> deregulation are logically sound.
>
> If economics can be brought back to earth and treated as a tool kit with
> which to analyse specific problems and propose appropriate solutions it
will
> be less in need of debunking. The world suffers enough from rigid
> ideologies and their practitioners, and would continue to do so if
economics
> was debunked out of existence. At the moment mainstream economics is part
> of the global problem, not a contributor to its solution,

I'm sure Robert just hasn't gotten around the Hayek-L yet, and his rebuke
will shortly be posted.

Patrick

Mark Patrick Witte

unread,
Mar 24, 2002, 8:09:55 PM3/24/02
to
Robert Vienneau, growing ever stranger, ever sadder.

William Hummel wrote:

H| "I would be interested to see an example of an economic
H| trend based purely on theory that has been validated
H| empirically"

Following up on earlier remarks in this exchange, I replied with:

W| "Well, take a look at Solow's work...or the paper I

W| recommended by Christiano." A nice place to read about
W| empirical tests of Solow is Charles I. Jones's book
W| _Introduction to Economic Growth_.

Robert Vienneau then entered the fray by editing my
response down to just the first seven words and then
claiming that I had not answered Mr. Hummel since
Solow's original papers we essentially work of theory
and contained little in the way of empirical support.
The key to my response was the citation of the book by
Charles Jones that summarizes a number of empirical
results that are supportive of Solow's model.

When this deceit of Mr. Vienneau's was laid bare, he
tries to hide is dishonesty by citing other papers on the
topic. There are two typically sad and strange Vienneau
aspects to this move.

1) Occasionally Robert Vienneau adopts the pretense
that he has read Thomas Kuhn's most famous book and to have
understood at least some parts important parts of what Kuhn
was trying to communicate. This is yet another striking
disproof of at least the latter part of such contentions.
In any research endeavor, there will be counter evidence
and neither Jones nor I nor anyone serious about this topic
would consider the matter settled. My citation of Jones
and his recent work should not be considered as closing the
book on the questions but was just what I said it was,
empirical support for Solow's theoretical growth work.

2) Mr. Vienneau's continuing citation of things he has
never read in support of his case, based solely upon the
title. Is the paper any good? Does it truly matter for
the question at hand? Who would know? Certainly not Mr.
Vienneau who has never read it. Do any of the papers he
cites, if he has read them or not, present alternative
empirical implementations that do a better job of
explaining the observed world than does the Solow approach?
If not, if they are just whining about alternative assumption
sets that have no superior support from data from the real world,
then consider this aspect of the thread ended.

In article <rvien-993721....@news.dreamscape.com>,

Mark Patrick Witte

unread,
Mar 24, 2002, 9:54:39 PM3/24/02
to
In article <jh0s9u0sdfv4qlmfs...@4ax.com>,

William F Hummel <wfhu...@attbi.com> wrote:
>On 24 Mar 2002 05:26:56 GMT, mwi...@merle.acns.nwu.edu (Mark
>Patrick Witte) wrote:
>
>> Mr. Hummel seems to have borrowed the aggressive editor from someone
>>else who posts on this thread and has left a number of questions I posed to
>>him unanswered. I'll try to answer his while recalling some of my previous
>>points, and be brief to the limits of my small ability on that score.
>>
>> * Does Mr. Hummel think that Dr. Donald Kohn is a fellow traveler
>etc. etc.
>
>Mr. Witte does love to ramble on. I've already stated my views
>on the letter from Kohn and have no intention of belaboring the
>issue. Answering a series of mostly rhetorical questions adds
>nothing to this discussion.

Cut to the chase: Do you think that Dr. Donald Kohn has a view of
how monetary policy works that is at odds with that of economists such as
Alan Blinder and Rich Mishkin? If so, is your who castle built upon the
sand of part of a letter from Kohn to Mosler?

>> * Does Mr. Hummel agree that the Federal reserve has made a
>>conscious policy *choice* to manage the Fed Funds Rate and it could instead
>>*choose* to manage reserves? I suspect that everyone on sci.econ (with one
>>possible exception) agrees that this is a good policy, but it is a policy
>>choice, not a law of nature.
>
>If Mr. Witte understood some of my earlier remarks about the
>Fed's actions in the early 1980s, he would not have asked this
>question. At least I would hope that's true.

I would hope that you would answer the question. Is the Feds
targeting the FFR a policy *choice* it has chosen? Could it (perhaps
imprudently) make a different policy *choice*?

>> * Does Mr. Hummel agree that if A + B = 10, and if we choose A to
>>be 7, then B must be 3? That is, having chosen A, we are left with a
>>Hobson's choice for B? And if we were instead to choose B first, then we
>>would have also at that moment limited our choice for A to a single number?
>>That is, that A and B are co-determined in the specified relationship?
>
>Of course, but this trivial example bears no relation to the
>question of whether the Fed can simultaneously control the Fed
>funds rate and the total amount of reserves as Mr. Witte has
>claimed, based on his reading of introductory macro textbooks.
>In fact it should be apparent that the two monetary variables are
>not dimensionally additive. I'll have more to say about how the
>Fed manages to control the overnight funds rate at another time.

It should be apparent that this is not an example of a monetary
model, but rather the simplest example of a co-determined relationship
available. And it begs the question Mr. Hummel raises, can the Fed
simultaneously and and independently control both rates and reserves?
Can it? (Hint: No, because these are co-determined.)

>> And now, I shall do my part:
>>
>>>Please answer a simple question. Let's assume banks issue loans
>>>this month well in excess of the aggregate reserves required to
>>>back them. (It can happen.) Assume also the FFR target has been
>>>selected previously and is required to be accurately met by the
>>>NYFRB. Tell me how your static model of X and Y are consistent
>>>with this imbalance, or what happens to restore equilibrium.
>>
>> As taught in Introductory Macroeconomics, since the Fed has
>>chosen to target the FFR, it will adjust the amount of reserves in
>>the system, either through expansionary OMOs, or through increased discount
>>lending. However, if the Fed where to instead target reserves, then
>>it would be the FFR that would bear the brunt of the adjustment, since these
>>variables are co-determined. However, in the real world, in either case,
>>the Fed would allow both variables to adjust to meet the demand shock for
>>reserves; that is, the FFR displays volitility around its target range and
>>the stock of private sector bank reserves has an elasticity that is
>>certainly not zero since the amount of bank reserves is less than 10% of the
>>available monetary base.
>
>Mr. Witte says that according to his reading of introductory
>macroeconomics, the Fed controls either the FFR or reserves and
>lets the other one adjust. [I would note here that is an example
>of a control variable and a dependent variable.] Then he
>modifies this by saying in the real world both variables adjust

Jeez, it it really so hard? Right now, the Fed would choose to
allow both variables to move to handle the shock, but in a co-determined
manner. The rates and the stock of reserves are not independent.

>together to meet the demand shock for reserves, although he
>doesn't acknowledge that one is the target variable and the other
>is a "don't care" variable. Apparently what is most important to
>him is that this must be seen as "co-determination".

No, the most important thing here is that they actually *are*
co-determined.

>Now it is true that the Fed uses its control of reserves to steer
>the Fed funds rate to its target value, a procedure I have
>described in some detail on sci.econ. But if that is the whole
>point, it is hardly worth the time and effort to say it. That's
>a case of not seeing the forest for the trees.

Indeed! The forest versus trees saying has been in my mind from the
beginning of this.

>Or to put it more
>directly, one cannot acquire a real understanding of how our
>monetary system works without seeing the bigger relationships.

Indeed! One cannot understand monetary policy, unless one realizes
the "bigger relationship" that the FFR and monetary base are co-determined,
as explained by myself, Blinder, and a host of others. The choice the
Fed correctly makes to target the FFR and adjust reserves accordingly is the
result of this relationship, and the choice in the past by the Fed and in
other nations to chose a policy of targeting reserves would leave the
adjustment to fall more heavily upon interest rates is a similar result of
this co-determination relationship.

So, is it agreed, that the FFR and monetary base are co-determined,
that if the Fed choses one, it then has no choice about the other?

Mr. Hummel suggests that "it is hardly worth the time and effort to
say it", but as I will explain in my next post, his earlier remarks make it
clear, this over-long debate was necessary.

>
>WFH

Robert Vienneau

unread,
Mar 25, 2002, 12:11:55 AM3/25/02
to
Mark Patrick Witte continues being silly:

> William Hummel wrote:
>
> H| "I would be interested to see an example of an economic
> H| trend based purely on theory that has been validated
> H| empirically"
>
> Following up on earlier remarks in this exchange, I replied with:
>
> W| "Well, take a look at Solow's work...or the paper I
> W| recommended by Christiano." A nice place to read about
> W| empirical tests of Solow is Charles I. Jones's book
> W| _Introduction to Economic Growth_.
>
> Robert Vienneau then entered the fray by editing my
> response down to just the first seven words and then
> claiming that I had not answered Mr. Hummel since
> Solow's original papers we essentially work of theory
> and contained little in the way of empirical support.

Saying that I claimed Solow's original papers were "essentially
work of theory" is, of course, to make an untrue statement. Solow's
original papers interpret empirical data, but contain little in the
way of theory, as I have explained repeatedly. Does Witte explain why


he disagrees with Samuelson's characterization of this work as like
that of a "busman on holiday who operates brilliantly and without

inhibitions in the rough-and-ready realm of empirical heuristics"? Does
Witte explain why he disagrees with Solow:

"I have never thought of the macroeconomic production function
as a rigorous justifiable concept. In my mind, it is either an
illuminating parable, or else a mere device for handling data, to
be used so long as it gives good empirical results, and to be
abandoned as soon as it doesn t, or as soon as something better
comes along."
-- Robert Solow, as quoted by Felipe and Fisher
<http://growthconf.ec.unipi.it/papers/Felipe.pdf>

(Section 10 of this paper is a summary of the main conclusions and
lessons. I like the metaphor that Witte and his ilk cannot bring
mermaids into existence merely by insisting on studying them.)

Witte, of course, does not answer these questions. Nor will
he address Felipe and Fisher:

"Aggregate production functions do not have a sound theoretical
foundation. For practical purposes this means that while generating
GDP, for example, as the sum of the components of aggregate demand
(or through the production or income sides of the economy) is
correct, generating and thinking of GDP as GDP=F(K,L), where K and
L are aggregates of capital and labor, respectively, is incorrect.
And thinking of aggregate investment as a well-defined addition to
capital in production is a mistake. The paper evaluates the standard
reasons given by economists for continuing to use aggregate production
functions in theoretical and applied work, and concludes that none of
them provides a valid argument."
-- Felipe and Fisher, 2001.

So I don't why Witte cites a heuristic for decomposing the
rate of growth as a "trend based purely on theory". I can see that
he seems to think that he can make a point by continuing to
misrepresent the opinion of others.



> The key to my response was the citation of the book by
> Charles Jones that summarizes a number of empirical
> results that are supportive of Solow's model.

I had written:

I don't know if the other references Mr. Witte gave address Franklin
Fisher's simulation results or contrast the empirical strength of
Solow's theory with competing theories (e.g., Kaldor's).

I still don't. I do know that Mr. Witte, for reasons that only can
be clear to himself, thinks raising these question about the
references he cites (other than Solow's) deceitful:

> When this deceit of Mr. Vienneau's was laid bare, he
> tries to hide is dishonesty by citing other papers on the
> topic.

And on and on and still on with boring and ignorant ad hominem.

I still don't know if Witte can cite any work that shows Solow's
approach to growth theory is empirically better than the
alternatives. Or, even, if the work Witte cites is just a matter
of estimating accounting identities. (That's rather the point of
Fisher's simulations and work by, e.g., McCombie, extending
Fisher's results.)

William F Hummel

unread,
Mar 25, 2002, 12:57:19 AM3/25/02
to
On 25 Mar 2002 02:54:39 GMT, mwi...@merle.acns.nwu.edu (Mark
Patrick Witte) wrote:

>In article <jh0s9u0sdfv4qlmfs...@4ax.com>,
>William F Hummel <wfhu...@attbi.com> wrote:
>>On 24 Mar 2002 05:26:56 GMT, mwi...@merle.acns.nwu.edu (Mark
>>Patrick Witte) wrote:
>>
>>> Mr. Hummel seems to have borrowed the aggressive editor from someone
>>>else who posts on this thread and has left a number of questions I posed to
>>>him unanswered. I'll try to answer his while recalling some of my previous
>>>points, and be brief to the limits of my small ability on that score.
>>>
>>> * Does Mr. Hummel think that Dr. Donald Kohn is a fellow traveler
>>etc. etc.
>>
>>Mr. Witte does love to ramble on. I've already stated my views
>>on the letter from Kohn and have no intention of belaboring the
>>issue. Answering a series of mostly rhetorical questions adds
>>nothing to this discussion.
>
> Cut to the chase: Do you think that Dr. Donald Kohn has a view of
>how monetary policy works that is at odds with that of economists such as
>Alan Blinder and Rich Mishkin?

Ambiguous question. How monetary policy works on the economy or
how monetary policy is implemented? If either case, the answer
is yes. I have never encountered any two people, economists or
otherwise, who fully agree on how the monetary system works.
Also I would guess Kohn knows far more about it than Blinder.
Kohn has been a specialist in the field for his whole career.
Sure, Blinder spent a year or so on the BOG, but it takes more
time than that to become an expert.

>If so, is your who castle built upon the
>sand of part of a letter from Kohn to Mosler?

Presumptuous question. No


>
>>> * Does Mr. Hummel agree that the Federal reserve has made a
>>>conscious policy *choice* to manage the Fed Funds Rate and it could instead
>>>*choose* to manage reserves? I suspect that everyone on sci.econ (with one
>>>possible exception) agrees that this is a good policy, but it is a policy
>>>choice, not a law of nature.
>>
>>If Mr. Witte understood some of my earlier remarks about the
>>Fed's actions in the early 1980s, he would not have asked this
>>question. At least I would hope that's true.
>
> I would hope that you would answer the question. Is the Feds
>targeting the FFR a policy *choice* it has chosen?

Of course. Fed monetary management methods have gone through
many variations over the years as the financial scene in the US
have evolved, particularly with the end of caps on interest
rates. An excellent review of this can be found in Marcia
Stigum's "Money Market", 3rd Edition. For at least the last 12
years, the Fed has tightly pegged the FFR, and it has explicitly
announced its target FFR since 1995. In the Fed's history you
won't find a longer period where the implementation has been
virtually unchanged.

>Could it (perhaps
>imprudently) make a different policy *choice*?

Yes, if by policy choice you mean its implementation. But the
current system of pegging the FFR works well, and in my opinion
it is a better choice than any of the systems the Fed has
previously tried.


>
>>> * Does Mr. Hummel agree that if A + B = 10, and if we choose A to
>>>be 7, then B must be 3? That is, having chosen A, we are left with a
>>>Hobson's choice for B? And if we were instead to choose B first, then we
>>>would have also at that moment limited our choice for A to a single number?
>>>That is, that A and B are co-determined in the specified relationship?
>>
>>Of course, but this trivial example bears no relation to the
>>question of whether the Fed can simultaneously control the Fed
>>funds rate and the total amount of reserves as Mr. Witte has
>>claimed, based on his reading of introductory macro textbooks.
>>In fact it should be apparent that the two monetary variables are
>>not dimensionally additive. I'll have more to say about how the
>>Fed manages to control the overnight funds rate at another time.
>
> It should be apparent that this is not an example of a monetary
>model, but rather the simplest example of a co-determined relationship
>available. And it begs the question Mr. Hummel raises, can the Fed
>simultaneously and and independently control both rates and reserves?
>Can it? (Hint: No, because these are co-determined.)

Then you agree that the quantity of reserves is a residual of the
pegging of the FFR?

Never said they were. When y = f(x), then x and y are obviously
not independent. Yet when x is taken as the independent variable
then y is the dependent variable. Call them co-determined if you
are determined to do so.


>
>>together to meet the demand shock for reserves, although he
>>doesn't acknowledge that one is the target variable and the other
>>is a "don't care" variable. Apparently what is most important to
>>him is that this must be seen as "co-determination".
>
> No, the most important thing here is that they actually *are*
>co-determined.

Ah yes, it's really very important to recognize that they are
co-determined. That really helps to see the big picture, right?


>
>>Now it is true that the Fed uses its control of reserves to steer
>>the Fed funds rate to its target value, a procedure I have
>>described in some detail on sci.econ. But if that is the whole
>>point, it is hardly worth the time and effort to say it. That's
>>a case of not seeing the forest for the trees.
>
> Indeed! The forest versus trees saying has been in my mind from the
>beginning of this.

Maybe you need to take aspirin.


>
>>Or to put it more
>>directly, one cannot acquire a real understanding of how our
>>monetary system works without seeing the bigger relationships.
>
> Indeed! One cannot understand monetary policy, unless one realizes
>the "bigger relationship" that the FFR and monetary base are co-determined,
>as explained by myself, Blinder, and a host of others.

Including the authors of introductory macro textbooks?

>The choice the
>Fed correctly makes to target the FFR and adjust reserves accordingly is the
>result of this relationship, and the choice in the past by the Fed and in
>other nations to chose a policy of targeting reserves would leave the
>adjustment to fall more heavily upon interest rates is a similar result of
>this co-determination relationship.

So the policy decision to target the FFR is the result of the
"co-determination" relationship? Hardly. It's because the FFR
is important in the macroeconomic sense, while reserves are at
least one order of magnitude down on that scale. Very few firms,
even banks, know (or care) what the total quantity of reserves
are, but they sure respond to changes in the FFR.

I can't understand why you insist on making so much of the fact
that the FFR and reserves are functionally related. You can't
change one without affecting the other, but so what? The
important point is that the Fed currently treats the FFR as the
target variable and the reserves as a "don't care" variable.
That is, the Fed doesn't really care what quantity of reserves it
must provide to peg the FFR at a chosen target. In that sense
reserves are the residual variable.

>
> So, is it agreed, that the FFR and monetary base are co-determined,
>that if the Fed choses one, it then has no choice about the other?

Absolutely not. The monetary base includes cash held by the
public, most of which is overseas according to the Fed. The
monetary base is an order of magnitude larger than the reserves ,
but it is the latter that banks respond to in bidding on fed
funds. Furthermore the FFR is a function of several variables,
not just aggregate reserves.


>
> Mr. Hummel suggests that "it is hardly worth the time and effort to
>say it", but as I will explain in my next post, his earlier remarks make it
>clear, this over-long debate was necessary.
>

Necessary for whom? Anyhow, don't forget to cite a lot of
references, and be sure to add footnotes. It won't sell if it
doesn't look authoritative.

WFH

Mark Patrick Witte

unread,
Mar 25, 2002, 3:14:58 AM3/25/02
to
Mr. Hummel seems to wonder, what is all the fuss about?
In his last post he wrote:

H| Now it is true that the Fed uses its control of
H| reserves to steer the Fed funds rate to its target
H| value, a procedure I have described in some detail on
H| sci.econ. But if that is the whole point, it is hardly
H| worth the time and effort to say it. That's a case of
H| not seeing the forest for the trees. Or to put it more
H| directly, one cannot acquire a real understanding of how
H| our monetary system works without seeing the bigger
H| relationships.

This begs two types of questions:

1) Why might it be "worth the time and effort to say it?"

Why? Because William Hummel has been using this point
for years as a proof of where much of the field of economics
gets something that he feels is important fundamentally wrong.

2) Who is "seeing the bigger relationships?"
Is it economists such as Friedman, Blinder, Christiano, Mishkin,
and the like? Or is it William Hummel? Are interest rates and
monetary reserves co-determined as Tinbergen explained in 1952
or does the Fed lack any control over reserves because it *must*
manage the interest rate?

How does what Mr. Hummel says above differ from
anything that I've said in these threads, or differ from the
piece by Alan Blinder that Mr. Hummel holds up as a signal
of basic misunderstanding by an well known economist?
http://groups.google.com/groups?q=Blinder+instrument+group:sci.econ+author:Hummel&hl=en&selm=cbdn9uorb06vmqmj3jq47iimojckrch4vt%404ax.com&rnum=1

(Blinder is "incorrect in some respects", but what can we expect? He was
only at the Fed for a couple of years, so how much could he really have
learned about monetary policy? What? People are chosen for Fed jobs based
upon their previous related accomplishments? Oh, then never mind.)

How does Mr. Hummel's statement differ from what is taught in those
terrible economics textbooks that he so derides (but never references)?

"When the demand curve for money shifts outward, the Fed
must tolerate a rise in interest rates, a rise in the money stock,
or both. It simply cannot control *both* the supply of money
*and* the interest rate. If it tries to keep M steady, then r will rise
sharply. Conversely, if it tries to stabilize r, then M will shoot up."
(Emphasis from the original, p. 286 of _Macroeconomics:
Principles and Policy_, Baumol and Blinder, 1999.)

Why does all this matter? It matters because Mr. Hummel's
misunderstanding of the operating choices of monetary authorities has
led him to believe and repeatedly state that central banks cannot
increase the supply of reserves in the system. It has led him to badly
mis-read Phillip Cagan. It has led him to disparage macroeconomists
in general because they don't share his misperception and so in his
eyes must lack a basic clue as to how the real world works.

It's a slightly subtle point, but one that has led Mr. Hummel into
some serious misunderstanding. It's a good thing that the Fed targets the
FFR, but it doesn't have to do so, and even targeting the FFR can lead to
problems if it's done at the wrong level, thus causing too much growth of
the monetary aggregates and resulting rising inflation.


In article <ptet9u0j31clicsa5...@4ax.com>,


William F Hummel <wfhu...@attbi.com> wrote:
>On 25 Mar 2002 02:54:39 GMT, mwi...@merle.acns.nwu.edu (Mark
>Patrick Witte) wrote:
>>
>> Cut to the chase: Do you think that Dr. Donald Kohn has a view of
>>how monetary policy works that is at odds with that of economists such as

>>Alan Blinder and Rick Mishkin?

>
>Ambiguous question. How monetary policy works on the economy or
>how monetary policy is implemented? If either case, the answer
>is yes. I have never encountered any two people, economists or
>otherwise, who fully agree on how the monetary system works.

Dodge attempt unsuccessful. Blinder and Mishkin may differ slightly in
their views on monetary policy; would you say that Kohn differs to a much
larger extent from these two guys than they differ from each other? And do
you base your answer from that single paragraph from a letter that was
responding to points unclear?

>Also I would guess Kohn knows far more about it than Blinder.
>Kohn has been a specialist in the field for his whole career.
>Sure, Blinder spent a year or so on the BOG, but it takes more
>time than that to become an expert.

Indeed, imagine Blinder's surprise when he hatched out of his egg
and found himself the Fed Vice-Chairman. Maybe it would have been better if
they had chosen someone who had spent his life studying and researching
monetary and macroeconomic issues and finding original insights into the
function of policy.

>>>> * Does Mr. Hummel agree that the Federal reserve has made a
>>>>conscious policy *choice* to manage the Fed Funds Rate and it could instead
>>>>*choose* to manage reserves? I suspect that everyone on sci.econ (with one
>>>>possible exception) agrees that this is a good policy, but it is a policy
>>>>choice, not a law of nature.
>>>
>>>If Mr. Witte understood some of my earlier remarks about the
>>>Fed's actions in the early 1980s, he would not have asked this
>>>question. At least I would hope that's true.
>>
>> I would hope that you would answer the question. Is the Feds
>>targeting the FFR a policy *choice* it has chosen?
>
>Of course. Fed monetary management methods have gone through

Yes! Yes! Yes!

>Stigum's "Money Market", 3rd Edition. For at least the last 12

Ah, a woman with great NU econ ties.

>>Could it (perhaps
>>imprudently) make a different policy *choice*?
>
>Yes, if by policy choice you mean its implementation. But the

By policy choice, I mean one of targeting reserves or some
aggregate, rather than the choice of targeting the interest
rate. I would hope that might become obvious after a hundred
repetitions or so.

>> It should be apparent that this is not an example of a monetary
>>model, but rather the simplest example of a co-determined relationship
>>available. And it begs the question Mr. Hummel raises, can the Fed
>>simultaneously and and independently control both rates and reserves?
>>Can it? (Hint: No, because these are co-determined.)
>
>Then you agree that the quantity of reserves is a residual of the
>pegging of the FFR?

As with the quote from Baumol and Blinder above, the two variables are
co-determined so that whichever one is chosen as the target relegates the
the other to do residual adjustment. Do a websearch on Google groups,
you'll find I've been saying just this for years.

Well, what exactly were you saying all these years?

(Oct 5 1998)
H| In the big picture, there is no question that the money supply
H| is endogenous and the Fed funds rate is exogenous.

(Dec 11, 2000)
H| The Fed cannot target the monetary base. The demand for
H| currency by the public must be met by the Fed.

(September 8, 2001)
H| There were and still are plenty of economists who have swallowed
H| his [Friedman's] views on the neutrality of money vis-a-vis the
H| real economy and the causal relation between money and inflation.

(March 19, 2002)
H| And for one who worked a spell at the Federal Reserve,
H| Mr. Witte seems unaware of the endogeneity of the base money
H| supply. His view on the cause of inflation is also surprising.

(March 19, 2002)
H| Not only is a causal relation impossible to show empirically,
H| but in this case it ignores the fact that the money supply is
H| endogenous.

(March 19, 2002)
H| Thus it makes little sense to assert that money growth is the
H| _cause_ of inflation, or even of hyperinflation.

(March 20, 2002)
H| The amount of fresh bank
H| lending responds (among other things) to the Fed funds
H| rate as expressed in bank lending rates.
H| I see nothing co-determined in that.

(March 20, 2002)
H| Yes, I meant Cagan. So you still conclude that
H| hyperinflations are CAUSED by printing too much money?

(March 21, 2002)
H| The notion that price and quantity of banking system
H| reserves are both controlled (co-determined?) by the
H| Fed is absurd.

(March 21, 2002)
H| Large banks often create more deposits through lending
H| than they have reserves for, and seek the reserves after
H| the fact. If this causes a shortage of aggregate reserves
H| in the banking system, the Fed must increase the supply
H| of reserves, typically through short term purchase
H| agreements with dealers, collateralized by Treasury
H| securities. The FOMC selects a target Fed funds rate,
H| and the NYFRB open market desk acts through its
H| OMO to hold Fed funds rate on target.

(March 21, 2002)
H| The Fed can only control the Fed funds rate.

(March 24, 2002)
(W| So, is it agreed, that the FFR and monetary base are co-determined,
that if the Fed choses one, it then has no choice about the other?)
H| Absolutely not.

With many of these quotes, broader context helps, but clearly
there is the suggestion that the dependence must run from quantities
to rates. However the Fed can chose to target either and so the
dependence is a choice, not a result of theory.

>>>together to meet the demand shock for reserves, although he
>>>doesn't acknowledge that one is the target variable and the other
>>>is a "don't care" variable. Apparently what is most important to
>>>him is that this must be seen as "co-determination".
>>
>> No, the most important thing here is that they actually *are*
>>co-determined.
>
>Ah yes, it's really very important to recognize that they are
>co-determined. That really helps to see the big picture, right?

Evidently. If one recognizes this, then the writings of Friedman,
Blinder, and Cagan make a lot of sense.

>> Indeed! One cannot understand monetary policy, unless one realizes
>>the "bigger relationship" that the FFR and monetary base are co-determined,
>>as explained by myself, Blinder, and a host of others.
>
>Including the authors of introductory macro textbooks?

Or all the non-intermediate books I listed before. Speaking of
lists of textbooks, where is yours of all the ones with terrible
inaccuracies in them?

>>The choice the
>>Fed correctly makes to target the FFR and adjust reserves accordingly is the
>>result of this relationship, and the choice in the past by the Fed and in
>>other nations to chose a policy of targeting reserves would leave the
>>adjustment to fall more heavily upon interest rates is a similar result of
>>this co-determination relationship.
>
>So the policy decision to target the FFR is the result of the
>"co-determination" relationship? Hardly.

Right, that is hardly what I said. Following Tinbergen, the decision of
monetary authories to target interest rates and support this target by
adjusting reserves is because of this co-determined relationship. It
recalls a really funny line about Jimmy Carter from William Greider's book.
(That is, _Secrets of the Temple_, Simon and Schuster, 1989).

>I can't understand why you insist on making so much of the fact
>that the FFR and reserves are functionally related. You can't
>change one without affecting the other, but so what? The
>important point is that the Fed currently treats the FFR as the
>target variable and the reserves as a "don't care" variable.
>That is, the Fed doesn't really care what quantity of reserves it
>must provide to peg the FFR at a chosen target. In that sense
>reserves are the residual variable.

This is exactly the point that I, Blinder, and a host of intro and
intermediate texts have been making for years. The point is that monetary
aggregates can be or not be endogenous depending upon the choice of the
monetary authority about what to target and what to adjust in compensation
to hit that target. Thus, when Friedman or Cagan write about exogenous
growth of the money supply, they are just examining a system where the
interest rate is not being set by the central bank. As such, they are not
wrong in the way that you have said Friedman is wrong.

>> So, is it agreed, that the FFR and monetary base are co-determined,
>>that if the Fed choses one, it then has no choice about the other?
>
>Absolutely not. The monetary base includes cash held by the
>public, most of which is overseas according to the Fed. The
>monetary base is an order of magnitude larger than the reserves ,
>but it is the latter that banks respond to in bidding on fed
>funds. Furthermore the FFR is a function of several variables,
>not just aggregate reserves.

Absolutely not? What choice does the Fed have about growth of the MB
given its FFR target?

>> Mr. Hummel suggests that "it is hardly worth the time and effort to
>>say it", but as I will explain in my next post, his earlier remarks make it
>>clear, this over-long debate was necessary.
>>
>Necessary for whom? Anyhow, don't forget to cite a lot of
>references, and be sure to add footnotes. It won't sell if it
>doesn't look authoritative.

Citing references in support of claims? What a concept!

Clearly it won't sell either way. I'm giving this away and have clearly
had no success selling books by authors like Kuhn and Blinder, even to
people who claim that they are interested in the topics.

Mark Patrick Witte

unread,
Mar 25, 2002, 3:21:35 AM3/25/02
to
In article <rvien-FF5BC4....@news.dreamscape.com>,

Robert Vienneau <rv...@see.sig.com> wrote:
>Mark Patrick Witte continues being silly:

Robert Vienneau recalls for me the saying, "There is no shaming
the shameless."

It's clear what he's trying to do, and it's ugly. If he feels
that models with assumptions that don't annoy his itchy bowel
syndrome have better empirical support than do Solow's, then
he is free to trot out that real world evidence. If he wants to
whine about modeling rigor and assumptions, he can take that
to the editors of JPE, QJE, JEP, AER, and Brookings who have
published all those papers in recent years by folks including
Jones, Barro, Sala-I-Martin, Romer, et al.

Concerning Vienneau's spineless behavior, let's be clear here.

William Hummel asked:
H| "I would be interested to see an example of an economic trend
H| based purely on theory that has been validated empirically"

In another context, I had recommended the work of Robert
Solow. To Mr. Hummel's reasonable further request from above,
I replied:


W| "Well, take a look at Solow's work...or the paper I
W| recommended by Christiano." A nice place to read about
W| empirical tests of Solow is Charles I. Jones's book
W|_Introduction to Economic Growth_.

Robert Vienneau edited my words to read:

W| "Well, take a look at Solow's work..."

and then Vienneau responded to them as though they were
my original statement.

Actually, it is even more dishonest than this. In response
to Mr. Hummel's misplaced request that I provide more
references, I offered the following extension:

W| Rather than have Mr. Hummel spend his time reading Solow's
W| 1956 QJE paper or his 1957 RE Stat paper, I directly
W| recommend: "A nice place to read about empirical tests of
W| Solow is Charles I. Jones's book _Introduction to Economic
W| Growth_." (Norton Press, 2nd edition, 2002).

In his same post, Mr. Vienneau edited this down to read
W| "Solow's 1956 QJE paper or his 1957 RE Stat paper..."

Mr. Vienneau then goes on to quote Solow as saying that
his 1957 paper "is in no sense a TEST of aggregate production
functions or marginal productivity or of anything else", as though I had
implied that it was. Mr. Vienneau follows this up with,

V| "Somewhat confusingly, Solow (1957) contains correlation
V| coefficients. So it seems Solow (1957) does not empirically
V| validate anything."

Mr. Vienneau has previously put on display his lack of
understanding of anything relating to empirical testing so what we
see here is consistent. There is nothing confusing about Solow's
1957 paper; in it Solow is applying the theory from his 1956 paper
to revolutionize the field of growth accounting. In his 1957 paper,
Solow works out his measure of total factor productivity as a
residual from effects of the embodied factors, creating what came
to be known as the "Solow residual." As such, this is an extension
of Solow's theory work, and not a direct test of his model, hence
my recommendation of work by Jones. It was dishonest of Mr.
Vienneau to imply that I was suggesting those Solow papers as direct
empirical verification of Solow's own work.

No reasonable person could read the Solow papers without
recognizing that they make up the theoretical framework that the
empirical tests worked out in Jones (and many other places) use. If
there are papers that Vienneau thinks work with better theories and
do a better job of explaining the real world data, he is free to bring
them out. In the end, that's what it's about, explaining the world
that we see. If Vienneau had some models that he liked that did a
good job of this, we would have seen them years ago. He doesn't,
so he whines about assumptions and mis-edits quotes.

William F Hummel

unread,
Mar 25, 2002, 12:19:54 PM3/25/02
to
On 25 Mar 2002 08:14:58 GMT, mwi...@merle.acns.nwu.edu (Mark
Patrick Witte) wrote:

> Well, what exactly were you saying all these years?

I owe Mr. Witte a thank you for taking the time and effort to
research what I have been saying all these years.

>
>(Oct 5 1998)
>H| In the big picture, there is no question that the money supply
>H| is endogenous and the Fed funds rate is exogenous.

Correct


>
>(Dec 11, 2000)
>H| The Fed cannot target the monetary base. The demand for
>H| currency by the public must be met by the Fed.

Correct


>
>(September 8, 2001)
>H| There were and still are plenty of economists who have swallowed
>H| his [Friedman's] views on the neutrality of money vis-a-vis the
>H| real economy and the causal relation between money and inflation.

Correct


>
>(March 19, 2002)
>H| And for one who worked a spell at the Federal Reserve,
>H| Mr. Witte seems unaware of the endogeneity of the base money
>H| supply. His view on the cause of inflation is also surprising.

Indeed


>
>(March 19, 2002)
>H| Not only is a causal relation impossible to show empirically,
>H| but in this case it ignores the fact that the money supply is
>H| endogenous.

Correct


>
>(March 19, 2002)
>H| Thus it makes little sense to assert that money growth is the
>H| _cause_ of inflation, or even of hyperinflation.
>

Yes, as Cagan has shown. Money growth is a symptom of a deeper
problem.

>(March 20, 2002)
>H| The amount of fresh bank
>H| lending responds (among other things) to the Fed funds
>H| rate as expressed in bank lending rates.
>H| I see nothing co-determined in that.

Correct.

>
>(March 20, 2002)
>H| Yes, I meant Cagan. So you still conclude that
>H| hyperinflations are CAUSED by printing too much money?

As above.


>
>(March 21, 2002)
>H| The notion that price and quantity of banking system
>H| reserves are both controlled (co-determined?) by the
>H| Fed is absurd.

Possibly quoted out of context. Obviously I mean the two cannot
simultaneously and independently be controlled by the Fed.


>
>(March 21, 2002)
>H| Large banks often create more deposits through lending
>H| than they have reserves for, and seek the reserves after
>H| the fact. If this causes a shortage of aggregate reserves
>H| in the banking system, the Fed must increase the supply
>H| of reserves, typically through short term purchase
>H| agreements with dealers, collateralized by Treasury
>H| securities. The FOMC selects a target Fed funds rate,
>H| and the NYFRB open market desk acts through its
>H| OMO to hold Fed funds rate on target.

Simple fact. Surprised to see it quoted here.


>
>(March 21, 2002)
>H| The Fed can only control the Fed funds rate.

If not out of context, this is incorrect.


>
>(March 24, 2002)
>(W| So, is it agreed, that the FFR and monetary base are co-determined,
>that if the Fed choses one, it then has no choice about the other?)
>H| Absolutely not.

Correct


>
> With many of these quotes, broader context helps, but clearly
>there is the suggestion that the dependence must run from quantities
>to rates.

True, but you did a fair job of avoiding out-of-context quotes.

>However the Fed can chose to target either and so the
>dependence is a choice, not a result of theory.

Great Scott! Do you mean this has been the bee in your bonnet
all the time? Where did I say the Fed cannot choose to target
reserves? I have pointed to the Volcker period when the Fed did
just that. My point, to reiterate, is that such an approach does
not provide a satisfactory result as the record clearly
indicates. That's why it is no longer done, and may never again
be done unless things really go wrong.


>
>>> No, the most important thing here is that they actually *are*
>>>co-determined.
>>
>>Ah yes, it's really very important to recognize that they are
>>co-determined. That really helps to see the big picture, right?
>
> Evidently. If one recognizes this, then the writings of Friedman,
>Blinder, and Cagan make a lot of sense.

As I have said many times, co-determination is a just fancy word
for functionally related. I have never claimed the FFR and
reserves were not functionally related, but you seem oblivious to
that fact.

>
>>> Indeed! One cannot understand monetary policy, unless one realizes
>>>the "bigger relationship" that the FFR and monetary base are co-determined,
>>>as explained by myself, Blinder, and a host of others.
>>
>>Including the authors of introductory macro textbooks?
>
> Or all the non-intermediate books I listed before. Speaking of
>lists of textbooks, where is yours of all the ones with terrible
>inaccuracies in them?

Mr. Witte does like to refer to textbooks in support of his
views.


>
>>>The choice the
>>>Fed correctly makes to target the FFR and adjust reserves accordingly is the
>>>result of this relationship, and the choice in the past by the Fed and in
>>>other nations to chose a policy of targeting reserves would leave the
>>>adjustment to fall more heavily upon interest rates is a similar result of
>>>this co-determination relationship.
>>
>>So the policy decision to target the FFR is the result of the
>>"co-determination" relationship? Hardly.
>
> Right, that is hardly what I said.

Then read it again. I think that is exactly what you said.

>Following Tinbergen, the decision of
>monetary authories to target interest rates and support this target by
>adjusting reserves is because of this co-determined relationship. It
>recalls a really funny line about Jimmy Carter from William Greider's book.
>(That is, _Secrets of the Temple_, Simon and Schuster, 1989).
>
>>I can't understand why you insist on making so much of the fact
>>that the FFR and reserves are functionally related. You can't
>>change one without affecting the other, but so what? The
>>important point is that the Fed currently treats the FFR as the
>>target variable and the reserves as a "don't care" variable.
>>That is, the Fed doesn't really care what quantity of reserves it
>>must provide to peg the FFR at a chosen target. In that sense
>>reserves are the residual variable.
>
> This is exactly the point that I, Blinder, and a host of intro and
>intermediate texts have been making for years.

If that is so, then what have you been carping about all this
time?

I note that there has been a subtle change in your reference from
introductory text books to intermediate text books. Have you had
a change of views about the quality of introductory text books?

>The point is that monetary
>aggregates can be or not be endogenous depending upon the choice of the
>monetary authority about what to target and what to adjust in compensation
>to hit that target.

In theory the monetary authority can control the monetary
aggregates, but it has never succeeded in doing so when it tried.
And it made a mess of things in the process. My interest in
theoretical possibilities diminishes sharply when they have
little or no use in practice. This is why I said it was very
misleading for Blinder to have said the Fed can control ANY
monetary aggregate. Of course as an academic he must cover all
the bases, or he would be criticized by his brethren.

>Thus, when Friedman or Cagan write about exogenous
>growth of the money supply, they are just examining a system where the
>interest rate is not being set by the central bank. As such, they are not
>wrong in the way that you have said Friedman is wrong.

What specifically did I say about Friedman that you think is
wrong?


>
>>> So, is it agreed, that the FFR and monetary base are co-determined,
>>>that if the Fed choses one, it then has no choice about the other?
>>
>>Absolutely not. The monetary base includes cash held by the
>>public, most of which is overseas according to the Fed. The
>>monetary base is an order of magnitude larger than the reserves ,
>>but it is the latter that banks respond to in bidding on fed
>>funds. Furthermore the FFR is a function of several variables,
>>not just aggregate reserves.
>
> Absolutely not? What choice does the Fed have about growth of the MB
>given its FFR target?

The MB can increase without affecting reserves when for example a
government agency makes direct cash purchases.

>
>>> Mr. Hummel suggests that "it is hardly worth the time and effort to
>>>say it", but as I will explain in my next post, his earlier remarks make it
>>>clear, this over-long debate was necessary.
>>>
>>Necessary for whom? Anyhow, don't forget to cite a lot of
>>references, and be sure to add footnotes. It won't sell if it
>>doesn't look authoritative.
>
> Citing references in support of claims? What a concept!

Ah yes, argue by appeal to "authority." That saves having to
develop one's own thoughts - hopefully.


>
> Clearly it won't sell either way. I'm giving this away and have clearly
>had no success selling books by authors like Kuhn and Blinder, even to
>people who claim that they are interested in the topics.

Perhaps you have been trying to hard.

WFH

Mark Patrick Witte

unread,
Mar 25, 2002, 2:44:33 PM3/25/02
to

Perhaps the motto of sci.econ should be, "Delete what you can't respond
to, but always respond." I've perhaps misallocated a great deal of time in
correspondence with Mr. Hummel under the apprehension that he was a bright
fellow of a scholarly bent who was serious about learning about how the
world works. Intelligent people can disagree but still learn from each
other and I had hopes. I perhaps need to re-evaluate.

The key exchange in the my last post was this:

H| I can't understand why you insist on making so much of the
H| fact that the FFR and reserves are functionally related. You can't
H| change one without affecting the other, but so what? The
H| important point is that the Fed currently treats the FFR as the
H| target variable and the reserves as a "don't care" variable.
H| That is, the Fed doesn't really care what quantity of reserves it
H| must provide to peg the FFR at a chosen target. In that sense
H| reserves are the residual variable.

W| This is exactly the point that I, Blinder, and a host of intro and
W| intermediate texts have been making for years. The point is that
W| monetary aggregates can be or not be endogenous depending upon the
W| choice of the monetary authority about what to target and what to adjust
W| in compensation to hit that target. Thus, when Friedman or Cagan write
W| about exogenous growth of the money supply, they are just examining a
W| system where the interest rate is not being set by the central bank. As
W| such, they are not wrong in the way that you have said Friedman is
W| wrong.

H| In theory the monetary authority can control the monetary
H| aggregates, but it has never succeeded in doing so when it tried.
H| And it made a mess of things in the process. My interest in
H| theoretical possibilities diminishes sharply when they have
H| little or no use in practice. This is why I said it was very
H| misleading for Blinder to have said the Fed can control ANY
H| monetary aggregate. Of course as an academic he must cover all
H| the bases, or he would be criticized by his brethren.

Here we have a classic "forest for the trees" misperception. It is
true that the financial system likes to have a stable discount rate,
but what is the really essential issue in monetary policy? Really?
It is to supply a sufficient money supply for the economy to grow
without inflation. Why does the Fed target the Fed Funds Rate, or
all things? Because that rate seems to give the quickest indication
of whether there is too much or too little reserves in the system.
Certainly the Fed values a stable financial system, but the financial
system is able to withstand shocks greater than even large
fluctuations in the interbank loan rate. The Fed manages the FFR
because it desires to head off recessions and inflations. To think
otherwise is to imagine that a physician who checks a pulse at the
wrist does so because he thinks the wrist is somehow vital to human
health. As Mr. Hummel has rightly pointed out, it is hard for a
monetary authority to know how much reserves to put into the system,
so the Fed has taken to watching the FFR as a guide. It is the
canary in the coalmine of monetary policy.

For some years, Mr. Hummel has held some strong views about the
failings of what he refers to as "mainstream" economics. Some of this
was based upon a general outsider's naiveti about how research is done,
but part of it was based upon what he saw as serious misunderstandings
of how monetary issues are approached. This has led him to make
condemning remarks about economists like Milton Friedman and Alan
Blinder who suggest that money growth can be exogenous and can lead
to rising inflation, and about the field of economics in general. (I
strongly
recommend that Mr. Hummel re-read Phillip Cagan's work before citing
him in support. Cagan treats money growth as exogenous and as the root
cause of the hyperinflations in the economies he studies.)

This has lead him to statements such as the following:

H| There were and still are plenty of economists who

H| have swallowed his [Friedman's] views on the neutrality of
H| money vis-a-vis the real economy and the causal relation
H| between money and inflation.

(On Friedman's statement about the causal relationship between money
growth and inflation.)
H| Assuming Mr. Witte believes this himself, I have to conclude
H that his devotion to mainstream economics is clouding his own vision.

When it comes to the field of economics in general, Mr. Hummel has said
things like:

H| One who finds monetary theory uninteresting and not worth
H| bothering with should not be dealing with macroeconomics.
H| Macro models that ignore the financial institutions of an economy
H| are pretty much worthless.

He has suggested there is a "tendency in mainstream macroeconomics to
treat monetary issues as having little relevance", as showing that serious
economists lack exposure to "the institutional side - the real world."
When I asked him where he comes by such a notion, he says:

H| Read any college macro text.

This brought us to a strange juncture. When Mr. Hummel asks for an
example of something, he complains if he feels he isn't given sufficient
references to find the work, which if true would be fair. However, having
given him such references, he now dismisses them as "arguing by appeal to
'authority.'" However, in the cases in these threads, references have been
given to back up explained claims, or unfortunately have not been given in
support of claims that needed them. In response to the above quote, I did a
quick survey of a number of leading introductory, intermediate, and
advanced undergraduate textbooks at my disposal, and found no real
evidence to support Mr. Hummel's statement. When I pressed him for
specifics on this point, he began what seemed to me to be some odd process
of making issue about whether I was referring to introductory or
intermediate books. He has never produced an example of a textbook that
fails his notion of what would be reasonable.

In what I hope again will be the penultimate post in this overlong
series, I'll ask again if Mr. Hummel has any objection to this issue
is treated in a standard introductory text. Is there anything wrong
with this?

"When the demand curve for money shifts outward, the Fed
must tolerate a rise in interest rates, a rise in the money stock,
or both. It simply cannot control *both* the supply of money
*and* the interest rate. If it tries to keep M steady, then r will rise
sharply. Conversely, if it tries to stabilize r, then M will shoot up."
(Emphasis from the original, p. 286 of _Macroeconomics:
Principles and Policy_, Baumol and Blinder, 1999.)


In article <jmmu9uornn77flvs3...@4ax.com>,


William F Hummel <wfhu...@attbi.com> wrote:

William F Hummel

unread,
Mar 25, 2002, 2:54:04 PM3/25/02
to
On 25 Mar 2002 08:14:58 GMT, mwi...@merle.acns.nwu.edu (Mark
Patrick Witte) wrote:

> Mr. Hummel seems to wonder, what is all the fuss about?
>In his last post he wrote:
>
>H| Now it is true that the Fed uses its control of
>H| reserves to steer the Fed funds rate to its target
>H| value, a procedure I have described in some detail on
>H| sci.econ. But if that is the whole point, it is hardly
>H| worth the time and effort to say it. That's a case of
>H| not seeing the forest for the trees. Or to put it more
>H| directly, one cannot acquire a real understanding of how
>H| our monetary system works without seeing the bigger
>H| relationships.
>
> This begs two types of questions:
>
> 1) Why might it be "worth the time and effort to say it?"
>
> Why? Because William Hummel has been using this point
>for years as a proof of where much of the field of economics
>gets something that he feels is important fundamentally wrong.

Mr. Witte has begun firing his shot gun in hopes of finally
hitting the target. His reference to "this point" needs
clarification. I wouldn't guess what it specifically refers to.


>
> 2) Who is "seeing the bigger relationships?"
>Is it economists such as Friedman, Blinder, Christiano, Mishkin,
>and the like? Or is it William Hummel? Are interest rates and
>monetary reserves co-determined as Tinbergen explained in 1952
>or does the Fed lack any control over reserves because it *must*
>manage the interest rate?

Mr. Witte is competing here for the credentialism prize. Argue
the case by citing well-known academics, presumed to have
superior knowledge to any layman. Ignore the fact that they
seldom fully agree among themselves. Then end by repeating once
again the same tired misinterpretation of my position, a practice
Mr. Witte has developed to a high art.


>
> How does what Mr. Hummel says above differ from
>anything that I've said in these threads, or differ from the
>piece by Alan Blinder that Mr. Hummel holds up as a signal
>of basic misunderstanding by an well known economist?
>http://groups.google.com/groups?q=Blinder+instrument+group:sci.econ+author:Hummel&hl=en&selm=cbdn9uorb06vmqmj3jq47iimojckrch4vt%404ax.com&rnum=1

Another reference to credentials. Argue by citing the views of a
well-known economist. It would be nice if Mr. Witte would
develop his own views and present the rationale in his own words.


>
>(Blinder is "incorrect in some respects", but what can we expect? He was
>only at the Fed for a couple of years, so how much could he really have
>learned about monetary policy? What? People are chosen for Fed jobs based
>upon their previous related accomplishments? Oh, then never mind.)

Still more credentialism. Blinder is smart, but not infallible.
The notion that people are chosen for Fed jobs based on prior
related accomplishments is not a universal rule. There have been
several incompetents and political hacks among them. I'm not
saying Blinder is one of those, but I don't hold in awe his views
on at least some monetary issues.


>
> How does Mr. Hummel's statement differ from what is taught in those
>terrible economics textbooks that he so derides (but never references)?

> "When the demand curve for money shifts outward, the Fed
>must tolerate a rise in interest rates, a rise in the money stock,
>or both. It simply cannot control *both* the supply of money
>*and* the interest rate. If it tries to keep M steady, then r will rise
>sharply. Conversely, if it tries to stabilize r, then M will shoot up."
>(Emphasis from the original, p. 286 of _Macroeconomics:
>Principles and Policy_, Baumol and Blinder, 1999.)

Mr. Witte has finally presented something substantial to discuss,
even though only by quoting one of his fellow economists. I
don't have Blinder's text, so there may be context missing. But
let's take the quote as a stand-alone statement. What's wrong
here is that neither the money supply nor the interest rate are
defined in measurable terms. Is M one of the monetary
aggregates? If so, which? Or is M really money in the sense of
the monetary base? Is r the overnight interest rate controlled
by the Fed or some nebulous market rate in the private sector.
This is a good example of what is wrong or misleading in
introductory texts. The words often seem plausible until we look
at them carefully.

If M refers to a monetary aggregate, then the statement should
better read the demand for "credit" rather than "money" because
it is really the banks the create most of the monetary
aggregates. Now the demand for credit is a function of a large
number of economic variables, only one of which is the price of
credit. Furthermore the price of short term credit is pretty
much controlled by the Fed. It is not a free variable as implied
in the Blinder's text.

The truth is that Blinder's "law" is simply not supported by the
evidence under almost any definition of the terms. An important
missing link is the role of the FOMC, a body of 12 people
(usually dominated by the chairman) that decide what the short
term rates will be for reasons that normally have little relation
to size of the money supply. Indeed the Fed has declared that it
no longer is officially targeting any monetary aggregate,
although everyone knows that has been true since the mid-1980s.

So you can mark this quoted passage down as an example of what is
wrong in mainstream macro texts.


>
> Why does all this matter? It matters because Mr. Hummel's
>misunderstanding of the operating choices of monetary authorities has
>led him to believe and repeatedly state that central banks cannot
>increase the supply of reserves in the system.

Mr. Witte continues to make this claim, but he should be
embarrassed to do. It is patently false.

>It has led him to badly mis-read Phillip Cagan.

Wrong. It is Mr. Witte who misread Cagan.

>It has led him to disparage macroeconomists
>in general because they don't share his misperception and so in his
>eyes must lack a basic clue as to how the real world works.

Silly comments like this are hardly worth a response. I would be
glad to provide a list of macroeconomists whose work I value, but
Mr. Witte probably wouldn't recognize most of the names.


>
> It's a slightly subtle point, but one that has led Mr. Hummel into
>some serious misunderstanding.

I suspect Mr. Witte's understanding of our monetary system is
good deal better for having read some of my posts. However I
would never expect him to say so.

>It's a good thing that the Fed targets the
>FFR, but it doesn't have to do so, and even targeting the FFR can lead to
>problems if it's done at the wrong level, thus causing too much growth of
>the monetary aggregates and resulting rising inflation.

This is the first valid point Mr. Witte has made in this post.


>
>In article <ptet9u0j31clicsa5...@4ax.com>,
>William F Hummel <wfhu...@attbi.com> wrote:
>>On 25 Mar 2002 02:54:39 GMT, mwi...@merle.acns.nwu.edu (Mark
>>Patrick Witte) wrote:
>>>
>>> Cut to the chase: Do you think that Dr. Donald Kohn has a view of
>>>how monetary policy works that is at odds with that of economists such as
>>>Alan Blinder and Rick Mishkin?
>>
>>Ambiguous question. How monetary policy works on the economy or
>>how monetary policy is implemented? If either case, the answer
>>is yes. I have never encountered any two people, economists or
>>otherwise, who fully agree on how the monetary system works.
>
> Dodge attempt unsuccessful. Blinder and Mishkin may differ slightly in
>their views on monetary policy; would you say that Kohn differs to a much
>larger extent from these two guys than they differ from each other? And do
>you base your answer from that single paragraph from a letter that was
>responding to points unclear?

I've already made my position clear on Kohn's letter.


>
>>Also I would guess Kohn knows far more about it than Blinder.
>>Kohn has been a specialist in the field for his whole career.
>>Sure, Blinder spent a year or so on the BOG, but it takes more
>>time than that to become an expert.
>
> Indeed, imagine Blinder's surprise when he hatched out of his egg
>and found himself the Fed Vice-Chairman. Maybe it would have been better if
>they had chosen someone who had spent his life studying and researching
>monetary and macroeconomic issues and finding original insights into the
>function of policy.

Ah yes, that Fed Vice-Chairman hatched and died not long
thereafter. Was it a natural death or was it murder?

WFH

William F Hummel

unread,
Mar 25, 2002, 4:28:48 PM3/25/02
to
On 25 Mar 2002 19:44:33 GMT, mwi...@merle.acns.nwu.edu (Mark

Well, let's clean up some of Mr. Witte's misconceptions. First,
the discount rate is not a free variable. It is set by the Fed
and can be kept fixed as long as the Fed chooses to. Second,
under the monetary policy regime of targeting the FFR, the
reserves in the system are precisely what are needed by the
banking system to support its lending, no more and no less.
Reserves must be added by the Fed when the banking system
increases its net lending. Otherwise the Fed could not keep the
FFR on target. The notion that there is a right amount of
reserves for the needs of the economy as Mr. Witte implies is a
striking example of his misunderstanding of how the monetary
system works.

>Certainly the Fed values a stable financial system, but the financial

>system is able to withstand shocks greater than even large
>fluctuations in the interbank loan rate. The Fed manages the FFR
>because it desires to head off recessions and inflations. To think
>otherwise is to imagine that a physician who checks a pulse at the
>wrist does so because he thinks the wrist is somehow vital to human
>health. As Mr. Hummel has rightly pointed out, it is hard for a
>monetary authority to know how much reserves to put into the system,
>so the Fed has taken to watching the FFR as a guide. It is the
>canary in the coal mine of monetary policy.

This compounds Mr. Witte's error noted above. If pegging the FFR
is the Fed's modus operandi, then only valid reason for the Fed
to adjust the reserves in the banking system is to control the
FFR, the benchmark for all short term interest rates. In any
case, firms do not care a whit about the amount of reserves in
the banking system. They do care very much about the cost of
borrowing. Just as Donald Kohn noted, "the causation runs more
from money [credit] to reserves than vice versa". In other
words, reserves will track the (credit) money supply and are
simply a residual in the monetary system.


>
> For some years, Mr. Hummel has held some strong views about the
>failings of what he refers to as "mainstream" economics. Some of this
>was based upon a general outsider's naiveti about how research is done,
>but part of it was based upon what he saw as serious misunderstandings
>of how monetary issues are approached. This has led him to make
>condemning remarks about economists like Milton Friedman and Alan
>Blinder who suggest that money growth can be exogenous and can lead
>to rising inflation, and about the field of economics in general. (I
>strongly
>recommend that Mr. Hummel re-read Phillip Cagan's work before citing
>him in support. Cagan treats money growth as exogenous and as the root
>cause of the hyperinflations in the economies he studies.)
>
> This has lead him to statements such as the following:
>
>H| There were and still are plenty of economists who
>H| have swallowed his [Friedman's] views on the neutrality of
>H| money vis-a-vis the real economy and the causal relation
>H| between money and inflation.
>

Would Mr. Witte actually deny the truth of this? There are even
some on sci.econ who have said as much.

>(On Friedman's statement about the causal relationship between money
>growth and inflation.)

>H| Assuming Mr. Witte believes this himself, I have to conclude
>H that his devotion to mainstream economics is clouding his own vision.

Perhaps I am not sugar-coating it sufficiently for Mr. Witte's
taste.


>
> When it comes to the field of economics in general, Mr. Hummel has said
>things like:
>
>H| One who finds monetary theory uninteresting and not worth
>H| bothering with should not be dealing with macroeconomics.
>H| Macro models that ignore the financial institutions of an economy
>H| are pretty much worthless.

With so much misunderstanding about the monetary system and the
vital role of its institutions in the economy at large, it's hard
to imagine how macroeconomists without a strong background in
that area can properly deal with such issues as employment,
inflation, and growth.


>
> He has suggested there is a "tendency in mainstream macroeconomics to
>treat monetary issues as having little relevance", as showing that serious
>economists lack exposure to "the institutional side - the real world."
>When I asked him where he comes by such a notion, he says:
>
>H| Read any college macro text.

Mr. Witte forgets how many times he has referred to the
authoritative content of introductory or intermediate text books.


>
> This brought us to a strange juncture. When Mr. Hummel asks for an
>example of something, he complains if he feels he isn't given sufficient
>references to find the work, which if true would be fair. However, having
>given him such references, he now dismisses them as "arguing by appeal to
>'authority.'"

I thought the jocular nature of my cal for references would not
have escaped Mr. Witte. Wasn't there also something about adding
footnotes?

The second point relates to Mr. Witte's tendency to cite
references more in the mode of credentialism than debate.

>However, in the cases in these threads, references have been
>given to back up explained claims, or unfortunately have not been given in
>support of claims that needed them. In response to the above quote, I did a
>quick survey of a number of leading introductory, intermediate, and
>advanced undergraduate textbooks at my disposal, and found no real
>evidence to support Mr. Hummel's statement. When I pressed him for
>specifics on this point, he began what seemed to me to be some odd process
>of making issue about whether I was referring to introductory or
>intermediate books. He has never produced an example of a textbook that
>fails his notion of what would be reasonable.

Just so happens I have. See my comments about the passage from
the Baumol and Blinder text in another post.


>
> In what I hope again will be the penultimate post in this overlong
>series, I'll ask again if Mr. Hummel has any objection to this issue
>is treated in a standard introductory text. Is there anything wrong
>with this?
>
> "When the demand curve for money shifts outward, the Fed
>must tolerate a rise in interest rates, a rise in the money stock,
>or both. It simply cannot control *both* the supply of money
>*and* the interest rate. If it tries to keep M steady, then r will rise
>sharply. Conversely, if it tries to stabilize r, then M will shoot up."
>(Emphasis from the original, p. 286 of _Macroeconomics:
>Principles and Policy_, Baumol and Blinder, 1999.)

Fortunate that Mr. Witte should point to this very quote. I
commented at some length in another post today on the objections
I have to it.

WFH

George

unread,
Mar 25, 2002, 5:14:34 PM3/25/02
to
someone wrote:

> The term "borrowed reserves" refers specifically to
>funds borrowed from the Fed, not from other financial
>institutions.
>

Great discussion . No wonder there are so many " happy skeptics" and
conspiracy buffs following banking - it is so complicated! I still feel that
borrowing reserves is bad. If a bank doesn't have enough reserves that's bad.


Mark Patrick Witte

unread,
Mar 25, 2002, 5:30:52 PM3/25/02
to
In article <b15v9uc2d2cbjvh35...@4ax.com>,

William F Hummel <wfhu...@attbi.com> wrote:

D'oh! Indeed I miss-typed, and it should be Fed Funds Rate.

>>but what is the really essential issue in monetary policy? Really?
>>It is to supply a sufficient money supply for the economy to grow
>>without inflation. Why does the Fed target the Fed Funds Rate, or
>>all things? Because that rate seems to give the quickest indication
>>of whether there is too much or too little reserves in the system.
>
>Well, let's clean up some of Mr. Witte's misconceptions. First,
>the discount rate is not a free variable. It is set by the Fed

>Second,


>under the monetary policy regime of targeting the FFR, the
>reserves in the system are precisely what are needed by the
>banking system to support its lending, no more and no less.
>Reserves must be added by the Fed when the banking system
>increases its net lending. Otherwise the Fed could not keep the
>FFR on target. The notion that there is a right amount of
>reserves for the needs of the economy as Mr. Witte implies is a
>striking example of his misunderstanding of how the monetary
>system works.

Rather it is a striking example of the way economics thinks about
marginal analysis. Phrases like "the reserves in the system are

precisely what are needed by the banking system to support its lending"

have no meaning without noting the conditional nature of lending and the
policy choice about the level of the targeted interest rate. It is this
short of thing that makes it unclear if Mr. Hummel understands the
implications of the relationship between choices of interest rates and
levels of reserves, which are...co-determined.

>>Certainly the Fed values a stable financial system, but the financial
>>system is able to withstand shocks greater than even large
>>fluctuations in the interbank loan rate. The Fed manages the FFR
>>because it desires to head off recessions and inflations. To think
>>otherwise is to imagine that a physician who checks a pulse at the
>>wrist does so because he thinks the wrist is somehow vital to human
>>health. As Mr. Hummel has rightly pointed out, it is hard for a
>>monetary authority to know how much reserves to put into the system,
>>so the Fed has taken to watching the FFR as a guide. It is the
>>canary in the coal mine of monetary policy.
>
>This compounds Mr. Witte's error noted above. If pegging the FFR
>is the Fed's modus operandi, then only valid reason for the Fed
>to adjust the reserves in the banking system is to control the
>FFR, the benchmark for all short term interest rates.

And why would the Fed want to change the FFR? To set the stock of
money (and interest rates) correctly to avoid recession or inflation.

>In any case, firms do not care a whit about the amount of reserves in
>the banking system. They do care very much about the cost of
>borrowing.

Yes, but the cost of borrowing is co-determined with the amount of
reserves in the system.

>Just as Donald Kohn noted, "the causation runs more
>from money [credit] to reserves than vice versa". In other
>words, reserves will track the (credit) money supply and are
>simply a residual in the monetary system.

"...However, the Federal Reserve can control the composition

of reserves between non-borrowed and borrowed reserves fairly
closely in the short run. By varying this mix the Federal
Reserve can influence the federal funds rate and the other
money market rates in the short run and thereby affect
growth of the money supply over a longer period."

So, the point is controlling the money supply over a longer
period...to avoid recessions and inflations. Indeed, Mr. Hummel says he has
always known that the Fed and either choose the FFR or a given level of
reserves and that one not targeted must adjust to accomodate the choice of
the first. However, the point of all of this is not that the Fed cares a
whit for either the FFR or the quantity of any aggregate but rather for
avoiding various bad macroeconomic outcomes. Hence, when economists like
Friedman and Blinder write about controling the stock of money, what they
are saying is not about day to day deviations any more than they would care
if the FFR misses its target by a few basis points. Instead, they are quite
reasonably talking about getting the co-determined interest rate/stock
variable(s) right to achieve the desired macroeconomic performance. This is
the forest.

>> For some years, Mr. Hummel has held some strong views about the
>>failings of what he refers to as "mainstream" economics. Some of this

>>was based upon a general outsider's naivete about how research is done,

>>but part of it was based upon what he saw as serious misunderstandings
>>of how monetary issues are approached. This has led him to make
>>condemning remarks about economists like Milton Friedman and Alan
>>Blinder who suggest that money growth can be exogenous and can lead
>>to rising inflation, and about the field of economics in general. (I
>>strongly
>>recommend that Mr. Hummel re-read Phillip Cagan's work before citing
>>him in support. Cagan treats money growth as exogenous and as the root
>>cause of the hyperinflations in the economies he studies.)
>>
>> This has lead him to statements such as the following:
>>
>>H| There were and still are plenty of economists who
>>H| have swallowed his [Friedman's] views on the neutrality of
>>H| money vis-a-vis the real economy and the causal relation
>>H| between money and inflation.
>>
>Would Mr. Witte actually deny the truth of this? There are even
>some on sci.econ who have said as much.

Friedman broadly makes the claim that excessive money growth (as he
carefully defines it) leads to inflation, and that money is neutral in the
long run but not the short run. Mr. Hummel should be careful that he does
not misunderstand what Friedman means by neutrality however. Mr. Hummel has
raised the valid point that losses in output caused by, for example,
monetary shocks, may not be made up in the long run. This is an interesting
research area that certainly contains some strong support for Mr. Hummel's
position. However, this does not relate to the concept of monetary
neutrality, but rather concerns the issue of "super-neutrality" where money
has no effect in either its level or its changing growth rates. Friedman's
point about neutrality is one of comparative levels: Would we be any better
off today if the money stock were 5% greater? Friedman would say no.

>>(On Friedman's statement about the causal relationship between money
>>growth and inflation.)
>>H| Assuming Mr. Witte believes this himself, I have to conclude
>>H that his devotion to mainstream economics is clouding his own vision.
>
>Perhaps I am not sugar-coating it sufficiently for Mr. Witte's
>taste.

Sugar coat this: excessive money growth causes rising inflation.

>> When it comes to the field of economics in general, Mr. Hummel has said
>>things like:
>>
>>H| One who finds monetary theory uninteresting and not worth
>>H| bothering with should not be dealing with macroeconomics.
>>H| Macro models that ignore the financial institutions of an economy
>>H| are pretty much worthless.
>
>With so much misunderstanding about the monetary system and the
>vital role of its institutions in the economy at large, it's hard
>to imagine how macroeconomists without a strong background in
>that area can properly deal with such issues as employment,
>inflation, and growth.

Inflation yes, growth and employment are not so hard to imagine.

>> He has suggested there is a "tendency in mainstream macroeconomics to
>>treat monetary issues as having little relevance", as showing that serious
>>economists lack exposure to "the institutional side - the real world."
>>When I asked him where he comes by such a notion, he says:
>>
>>H| Read any college macro text.
>
>Mr. Witte forgets how many times he has referred to the
>authoritative content of introductory or intermediate text books.

No, I refer to textbooks for their central role as a place for someone
with a weak background in some topic to find a careful explanation of the
material of interest. This is the role and comparative advantage of
textbooks for passing along information to people who are truly interested
in some subject, particularly relative to typed out newsgroup discussion.

However, note that it was Mr. Hummel who brought textbooks into this
discussion as support for his claims about the inadequacy of "mainstream
macroecomics", but he has never been able to produce a list of such
offending texts. As such, he used textbooks as a call to authority,
that is he offered no support for his position except to point to some
books, yet he is unable to say exactly what books he means. It's really
sad, given Mr. Hummel's other complaints about citations.

Robert Vienneau

unread,
Mar 25, 2002, 7:00:15 PM3/25/02
to
Mark Patrick Witte continues being silly:

> Robert Vienneau recalls for me the saying, "There is no shaming
> the shameless."

Ad hominem.



> It's clear what he's trying to do, and it's ugly. If he feels
> that models with assumptions that don't annoy his itchy bowel
> syndrome have better empirical support than do Solow's, then
> he is free to trot out that real world evidence.

Ad hominem.

Notice that Mr. Witte does not and will not address the lack of
theoretical foundation for, say, Solovian growth theory. For
example, he does not explain why he disagrees with Felipe and
Fisher. He changes the subject. But even then, his position
seems weak.

Notice that Mr. Witte does not claim to have found any literature
at all that shows Solow's approach is empirically superior to
alternatives. I, on the other hand, have occasionally cited
evidence that certain other approaches are empirically
superior to canonical neoclassical models in direct statistical
tests.

> If he wants to
> whine about modeling rigor and assumptions, he can take that
> to the editors of JPE, QJE, JEP, AER, and Brookings who have
> published all those papers in recent years by folks including
> Jones, Barro, Sala-I-Martin, Romer, et al.

Ad hominem, strawpersons, and argument from authority.

By the way, I notice that Mr. Witte has never told us about the
AEA's reception of the Schelling committee report.

One might conclude from arguments that Mr. Witte ignores:

"The revival of growth theory during the last two decades no doubt
has produced important discussions, and seemingly interesting
empirical results. However, the authors do not realize that they
are using a tool whose invalidity was demonstrated decades ago. The
consequence is that these empirical results are unjustifiable and
even misleading. For example, an important aspect emphasized by the
new models is the idea of increasing returns at the aggregate level.
However, the aggregate production functions derived theoretically
have constant returns to scale (recall that the aggregation conditions
strongly depend on the existence of constant returns to scale at the
micro level, and that with non-constant returns aggregates do not
exist in general).

We also have to make a reference to the important fact that the new
endogenous growth models have introduced as a factor of production
a very problematic concept, namely that of the physical quantity of
human capital. What are the logical foundations, or conditions under
which this can be represented?"
-- Felipe and Fisher
<http://growthconf.ec.unipi.it/papers/Felipe.pdf>


> Concerning Vienneau's spineless behavior, let's be clear here.

Ad hominem.



> William Hummel asked:
> H| "I would be interested to see an example of an economic trend
> H| based purely on theory that has been validated empirically"

> In another context, I had recommended the work of Robert
> Solow. To Mr. Hummel's reasonable further request from above,
> I replied:

> W| "Well, take a look at Solow's work...or the paper I
> W| recommended by Christiano." A nice place to read about
> W| empirical tests of Solow is Charles I. Jones's book
> W|_Introduction to Economic Growth_.

> Robert Vienneau edited my words to read:

> W| "Well, take a look at Solow's work..."

> and then Vienneau responded to them as though they were
> my original statement.

The claim seems to be that I tried to convey to the reader
the view that Mr. Witte had not mentioned any other references.
This is extremely odd. For one thing, our readers might remember


what they had read two days ago. For another thing, I wrote:

"I don't know if the other references Mr. Witte gave address Franklin
Fisher's simulation results or contrast the empirical strength of
Solow's theory with competing theories (e.g., Kaldor's)."

Perhaps the reason Mr. Witte turns the discussion to personalities
and responds with nothing substantial is because the answer
to these implied questions is "No". Or perhaps the problem is
simply that Mr. Witte has not grown up yet.

> [ More of the same boring nonsense. ]

> Mr. Vienneau then goes on to quote Solow as saying that
> his 1957 paper "is in no sense a TEST of aggregate production
> functions or marginal productivity or of anything else", as though I had
> implied that it was.

Strawperson.

Mr. Witte can draw any inferences he wants. In pointing out that the
first of his references does not empirically validate any trend
based purely on theory and wondering if his further references
do the same, I am not making any such implication.

> Mr. Vienneau follows this up with,

> V| "Somewhat confusingly, Solow (1957) contains correlation
> V| coefficients. So it seems Solow (1957) does not empirically
> V| validate anything."

> Mr. Vienneau has previously put on display his lack of
> understanding of anything relating to empirical testing so what we
> see here is consistent.

Ad hominem.

> There is nothing confusing about Solow's
> 1957 paper; in it Solow is applying the theory from his 1956 paper
> to revolutionize the field of growth accounting. In his 1957 paper,
> Solow works out his measure of total factor productivity as a
> residual from effects of the embodied factors, creating what came
> to be known as the "Solow residual." As such, this is an extension
> of Solow's theory work,

That's odd. Mr. Witte had written:

> > >Solow's original papers we essentially work of theory...

I responded with:

> >Saying that I claimed Solow's original papers were "essentially
> >work of theory" is, of course, to make an untrue statement. Solow's
> >original papers interpret empirical data, but contain little in the
> >way of theory, as I have explained repeatedly.

So above we see Mr. Witte agreeing with me somewhat - "applying the
theory" ('cept there isn't much theory there) - about the 1957
paper, but refusing to say so.

Attempts to provide theoretical foundation for Solow's work followed
these papers. As Samuelson, Burmeister, Fisher, and maybe even
Solow acknowledge, these attempts failed.

But I certainly want to point out that it is possible to write
papers with statistical tests and correlation coefficients - e.g.,
as in Solow - and not empirically test anything. We have more
agreement from Mr. Witte followed by an ad hominem illustrating
that he is too cowardly to say so or apologize for his silly
comments:

> and not a direct test of his model, hence
> my recommendation of work by Jones. It was dishonest of Mr.
> Vienneau to imply that I was suggesting those Solow papers as direct
> empirical verification of Solow's own work.

This "dishonesty" must be on display when I wrote:

"I don't know if the other references Mr. Witte gave address Franklin
Fisher's simulation results or contrast the empirical strength of
Solow's theory with competing theories (e.g., Kaldor's)."

Despite Mr. Witte's comments above, I still don't know if the other
references Mr. Witte gives address Franklin Fisher's simulation
results.


> No reasonable person could read the Solow papers without
> recognizing that they make up the theoretical framework that the
> empirical tests worked out in Jones (and many other places) use.

As Fisher knows, these empirical tests in many places are not
necessarily tests of the theory.

> If
> there are papers that Vienneau thinks work with better theories and
> do a better job of explaining the real world data, he is free to bring
> them out.

When Mr. Witte is referred to arguments about the lack of
theoretical foundation for the Solovian model, he changes the
subject.

> In the end, that's what it's about, explaining the world
> that we see.

That's odd. Mr. Witte recommends a paper that gives advise to the
U.S. congress on protectionism against Japanese imports based on
a study of mermaids. I don't see how life on this planet can be
explained like so.

> If Vienneau had some models that he liked that did a
> good job of this, we would have seen them years ago.

Felipe and Fisher are clear in their last paragraph that work I
have often cited provides an explaination of the world that we
see, an alternative to the failed paradigm Mr. Witte espouses.
(See also their 28th footnote.)

For some reason that cannot be clear to us, Mr. Witte is afraid
to acknowledge the existence of such work. So he must state
untruths about what I write instead:

> He doesn't,
> so he whines about assumptions and mis-edits quotes.

It seems the following would be simple questions to answer:

Does Mr. Witte give references that address Fisher's simulation
results and work extending Fisher's results?

Does Mr. Witte give references that contrast the empirical strength
of Solow's theory with competing theories (e.g., Kaldor's)?

William F Hummel

unread,
Mar 25, 2002, 8:02:34 PM3/25/02
to
On 25 Mar 2002 22:30:52 GMT, mwi...@merle.acns.nwu.edu (Mark
Patrick Witte) wrote:

>In article <b15v9uc2d2cbjvh35...@4ax.com>,
>William F Hummel <wfhu...@attbi.com> wrote:
>
>>Well, let's clean up some of Mr. Witte's misconceptions. First,
>>the discount rate is not a free variable. It is set by the Fed
>
>>Second,
>>under the monetary policy regime of targeting the FFR, the
>>reserves in the system are precisely what are needed by the
>>banking system to support its lending, no more and no less.
>>Reserves must be added by the Fed when the banking system
>>increases its net lending. Otherwise the Fed could not keep the
>>FFR on target. The notion that there is a right amount of
>>reserves for the needs of the economy as Mr. Witte implies is a
>>striking example of his misunderstanding of how the monetary
>>system works.
>
> Rather it is a striking example of the way economics thinks about
>marginal analysis. Phrases like "the reserves in the system are
>precisely what are needed by the banking system to support its lending"
>have no meaning without noting the conditional nature of lending and the
>policy choice about the level of the targeted interest rate. It is this
>short of thing that makes it unclear if Mr. Hummel understands the
>implications of the relationship between choices of interest rates and
>levels of reserves, which are...co-determined.

Mr. Witte can rest assured that I understand the relationships
involved here. I have reiterated them so many times in this
forum, it's surprising he raise the question, unless of course he
misunderstands them himself. In any case it should not be
necessary to write a treatise in support every statement one
makes.

>
>>This compounds Mr. Witte's error noted above. If pegging the FFR
>>is the Fed's modus operandi, then only valid reason for the Fed
>>to adjust the reserves in the banking system is to control the
>>FFR, the benchmark for all short term interest rates.
>
> And why would the Fed want to change the FFR? To set the stock of
>money (and interest rates) correctly to avoid recession or inflation.

A non-sequitur since the subject was about pegging the FFR, not
changing it. Nevertheless, I would hope Mr. Witte doesn't really
mean to imply it's possible for the Fed, by changing the FFR, to
"set the stock of money." The best it can hope to do is
influence the demand for credit over the longer run for the
purpose he correctly noted.

>
>>In any case, firms do not care a whit about the amount of reserves in
>>the banking system. They do care very much about the cost of
>>borrowing.
>
> Yes, but the cost of borrowing is co-determined with the amount of
>reserves in the system.

Mr. Witte loves that word "co-determined". It must have a
special meaning to him, or perhaps it just resonates well.


>
>>Just as Donald Kohn noted, "the causation runs more
>>from money [credit] to reserves than vice versa". In other
>>words, reserves will track the (credit) money supply and are
>>simply a residual in the monetary system.
>
> "...However, the Federal Reserve can control the composition
>of reserves between non-borrowed and borrowed reserves fairly
>closely in the short run. By varying this mix the Federal
>Reserve can influence the federal funds rate and the other
>money market rates in the short run and thereby affect
>growth of the money supply over a longer period."
>
> So, the point is controlling the money supply over a longer
>period...to avoid recessions and inflations.

Yes, the Fed is constantly working to avoid recessions and
inflations. But no, the Fed has no target for the money supply,
and without a target it's meaningless to talk about control. In
fact the Fed is on record as declaring that the relation between
any of the monetary aggregates and its primary policy objectives
is unpredictable, and has abandoned targeting the monetary
aggregates. The notion that the Fed controls the money supply is
one of the hoary errors of mainstream economics. It pervades the
textbooks, sometimes implicitly so the student is not even aware
of the assumption. It is one of the misconceptions I have been
trying to exorcise, an uphill battle in a forum where most have
been indoctrinated with mainstream training.

>Indeed, Mr. Hummel says he has
>always known that the Fed and either choose the FFR or a given level of
>reserves and that one not targeted must adjust to accomodate the choice of
>the first. However, the point of all of this is not that the Fed cares a
>whit for either the FFR or the quantity of any aggregate but rather for
>avoiding various bad macroeconomic outcomes.

The Fed talks about primary targets, and intermediate targets.
Obviously the primary targets are macroeconomic outcomes, but
that does mean the Fed cares not one whit about what it has to do
in the attempt to achieve those outcomes. Mr. Witte is playing
word games here.

>Hence, when economists like
>Friedman and Blinder write about controling the stock of money, what they
>are saying is not about day to day deviations any more than they would care
>if the FFR misses its target by a few basis points. Instead, they are quite
>reasonably talking about getting the co-determined interest rate/stock
>variable(s) right to achieve the desired macroeconomic performance. This is
>the forest.

Keep walking. You may yet find the way out.


>
> Friedman broadly makes the claim that excessive money growth (as he
>carefully defines it) leads to inflation, and that money is neutral in the
>long run but not the short run. Mr. Hummel should be careful that he does
>not misunderstand what Friedman means by neutrality however. Mr. Hummel has
>raised the valid point that losses in output caused by, for example,
>monetary shocks, may not be made up in the long run. This is an interesting
>research area that certainly contains some strong support for Mr. Hummel's
>position. However, this does not relate to the concept of monetary
>neutrality, but rather concerns the issue of "super-neutrality" where money
>has no effect in either its level or its changing growth rates. Friedman's
>point about neutrality is one of comparative levels: Would we be any better
>off today if the money stock were 5% greater? Friedman would say no.

Milton Friedman once declared "nothing is so unimportant as the
quantity of money expressed in terms of the nominal monetary
unit. . . The situation is very different with respect to the
real quantity of money."

Would you care to explain this in the context of the neutrality
of money?


>
>>Mr. Witte forgets how many times he has referred to the
>>authoritative content of introductory or intermediate text books.
>
> No, I refer to textbooks for their central role as a place for someone
>with a weak background in some topic to find a careful explanation of the
>material of interest. This is the role and comparative advantage of
>textbooks for passing along information to people who are truly interested
>in some subject, particularly relative to typed out newsgroup discussion.

Perhaps the central role of the introductory textbooks is the
main source of the problem in macroeconomics teaching today.


>
> However, note that it was Mr. Hummel who brought textbooks into this
>discussion as support for his claims about the inadequacy of "mainstream
>macroecomics", but he has never been able to produce a list of such
>offending texts. As such, he used textbooks as a call to authority,
>that is he offered no support for his position except to point to some
>books, yet he is unable to say exactly what books he means. It's really
>sad, given Mr. Hummel's other complaints about citations.

Wrong. You are ignoring the example I gave about the Blinder
textbook.

WFH

Mark Patrick Witte

unread,
Mar 25, 2002, 8:39:39 PM3/25/02
to
Let's see...what do we have here? Ah, insults, whining about
aggregation assumptions (yawn), insults, mermaids?, insults,...any empirical
support for approaches that Vienneau finds acceptable? Nope.

(That he would bring up the Schelling report is particularly amusing with
its call for more empirical work to be published. The critique of the
report is that there are a lot of competing journals that are always
competiting to attact readers by looking to publish papers that make
unusual claims and then back them up with sound theory and good
empirical work.)

In article <rvien-DAAF52....@news.dreamscape.com>,

Mark Patrick Witte

unread,
Mar 25, 2002, 8:57:37 PM3/25/02
to
In article <5phv9u8kek6h4vioj...@4ax.com>,

William F Hummel <wfhu...@attbi.com> wrote:
>On 25 Mar 2002 22:30:52 GMT, mwi...@merle.acns.nwu.edu (Mark
>Patrick Witte) wrote:
>
>> So, the point is controlling the money supply over a longer
>>period...to avoid recessions and inflations.
>
>Yes, the Fed is constantly working to avoid recessions and
>inflations. But no, the Fed has no target for the money supply,
>and without a target it's meaningless to talk about control. In
>fact the Fed is on record as declaring that the relation between
>any of the monetary aggregates and its primary policy objectives
>is unpredictable, and has abandoned targeting the monetary
>aggregates.

Indeed, because it is targeting the co-determined interest rate
variable. It should be noted that the relationship between the FFR and
other measures of economic activity is not stable either, hence its
frequent alteration.

>The notion that the Fed controls the money supply is
>one of the hoary errors of mainstream economics. It pervades the
>textbooks, sometimes implicitly so the student is not even aware
>of the assumption. It is one of the misconceptions I have been
>trying to exorcise, an uphill battle in a forum where most have
>been indoctrinated with mainstream training.

Classic. Remember that.

>Milton Friedman once declared "nothing is so unimportant as the
>quantity of money expressed in terms of the nominal monetary
>unit. . . The situation is very different with respect to the
>real quantity of money."
>
>Would you care to explain this in the context of the neutrality
>of money?

If the Fed had run a policy whereby the nominal money supply were
twice as high today as it is, prices would be twice as high, but the real
economy would be essentially unchanged. However, a change in the real
supply of money can bring on things like the Great Depression. Is
Friedman's intent so opaque here?

>>>Mr. Witte forgets how many times he has referred to the
>>>authoritative content of introductory or intermediate text books.
>>
>> No, I refer to textbooks for their central role as a place for someone
>>with a weak background in some topic to find a careful explanation of the
>>material of interest. This is the role and comparative advantage of
>>textbooks for passing along information to people who are truly interested
>>in some subject, particularly relative to typed out newsgroup discussion.
>
>Perhaps the central role of the introductory textbooks is the
>main source of the problem in macroeconomics teaching today.

Examples of those problems? What do texts get wrong that you get right?
Please tell me, I have a meeting with a rep from South-Western texts this
week. Please cite titles and problems.

>> However, note that it was Mr. Hummel who brought textbooks into this
>>discussion as support for his claims about the inadequacy of "mainstream
>>macroecomics", but he has never been able to produce a list of such
>>offending texts. As such, he used textbooks as a call to authority,
>>that is he offered no support for his position except to point to some
>>books, yet he is unable to say exactly what books he means. It's really
>>sad, given Mr. Hummel's other complaints about citations.
>
>Wrong. You are ignoring the example I gave about the Blinder
>textbook.

You gave an example from a Blinder textbook? No, I gave an example
from a Blinder textbook, one that you mis-interpreted as badly as you
seem to misinterpret Cagan. Please supply a some elements from your list of
"any college macro text that contains serious errors. And please give a
citation for what Cagan article you are talking about so that I can discuss
it as I will shortly do with your Blinder misunderstading.

Mark Patrick Witte

unread,
Mar 26, 2002, 12:06:20 AM3/26/02
to
In article <ifvu9u0atichsuek4...@4ax.com>,

William F Hummel <wfhu...@attbi.com> wrote:
>On 25 Mar 2002 08:14:58 GMT, mwi...@merle.acns.nwu.edu (Mark
>Patrick Witte) wrote:

>> Why does all this matter? It matters because Mr. Hummel's
>>misunderstanding of the operating choices of monetary authorities has
>>led him to believe and repeatedly state that central banks cannot
>>increase the supply of reserves in the system.
>
>Mr. Witte continues to make this claim, but he should be
>embarrassed to do. It is patently false.
>
>>It has led him to badly mis-read Phillip Cagan.
>
>Wrong. It is Mr. Witte who misread Cagan.

While I know it is credentialist of me to post or ask for citations,
I really would like to know what part of this classic old work on monetary
theory I am mis-reading. Earlier post a specific reference and quote from
a Cagan article that relates to real effects of exogenous increases in the
money supply, but if Mr. Hummel has in mind another piece of research by Mr.
Cagan, I hope that he will be good enough to give the reference and his
interpretation of the paper.

Mark Patrick Witte

unread,
Mar 26, 2002, 3:40:16 AM3/26/02