Re: Dollarization -- Moore's argument

["Colin Danby" <danbyc@xxxxxxx>]

Dear Basil,

Thank you for your reply.  Let me try to make the critical arguments
clearer.  I see three, though they are all at base the same point.

1. The labor mobility point, as raised recently by Paul -- that workers
cannot easily follow financial flows from surplus to deficit areas.  This
gets to the question of what kind of adjustment sequence is expected in the
deficit area.  It might be useful if you described the manner of adjustment
in the deficit area that you foresee with dollarization.

2. The Bagehot point: that a banking system needs a central bank that is
able to lend freely in the face of panic.  Lacking this capacity and given
fractional-reserve banking your system is vulnerable to runs, which are
essentially runs on the foreign assets held by the central bank.  Of course
this problem can already exist to a significant degree with a fixed exchange
rate plus capital mobility, as I have argued elsewhere.

On your specific responses:

> There will be no role for a local CB to set interest rates, but it still
> must oversee and exercise surveilance over the local banking system, re
> capital adequacy, etc,

The Bagehot argument is that this just is not enough.  You remain vulnerable
to runs.  (This is especially true in poorer, less internally-diverse, more
trade-dependent economies.)

> This should assure no extra lenders risk, and there will be no exchange
> rate risk if dollarization is credible.

The key word here is "credible," which I have argued over with Alan in the
past.  I think the term is typically used tautologically -- e.g. if your peg
collapses it wasn't credible.  Credible (or not) is an overbroad
*description* of the properties of the whole system; we need analysis.  Both
"credibility" and "country risk" function in the same merely-descriptive
way, as a means of separating off the messy stuff.

> The CB can still act as lender of
> last resort to provide system liquidity, as at present.

Only until its foreign reserves are exhausted.

> Think of a branch bank. Under dollarization we no longer have to ensure
> that each branch has no deficit or surplus, and depress the local economy
> if it should run a deficit.

Maybe I'm missing something obvious, but I don't get this.  The branch bank
example assumes a single banking system with banks (and obviously individual
branches) subject to the same kind of central bank support.  I am arguing
precisely that the international financial system is heterogeneous in ways
that prevent us assuming this.


3. The BoP/liquidity point: The definition of BoP "problems" you give sounds
like the familiar division of the BoP framework into autonomous and
accommodating segments, with the current account autonomous (trade and
various payments) and the capital account accommodating (borrowing and use
of reserve assets).  A "problem" is then when a deficit opens up in the
"autonomous" part that the "accommodating" part has enough difficulty
covering that some abrupt adjustment is required like a devaluation or
austerity plan.  Identifying when there is or is not a "problem" can
sometimes be rather subtle.  On the one hand govts may declare a "problem"
and invite in the IMF long before reserves actually vanish.  On the other
hand the very existence of a "problem" often depends on expectations of
various future flows and abilities to refinance plus expectations of
bailouts.  In any case it really comes down to distinguishing among good and
bad adjustment sequences, since the adjustment can’t not occur, ex post we
must balance.  Pkers I think would tend to see as problematic those
adjustments that involve large amounts of unemployment.

I note in passing, as I think Kindleberger argues, that even if the
distinction accommodating/autonomous is useful, different countries may have
totally different parts of the BoP in one or the other category.

But I am more interested in the question of where liquidity enters.  In the
textbook BoP scheme that autonomous/accommodating invites us to imagine, the
autonomous stuff is basically bartered with the outside world, and then the
plus or minus residual is settled in the accommodating category, by
borrowing or a transfer of reserve assets.

A PK analysis would suggest that liquidity enters in not so much to this
final settlement, but into the very ability to carry out commerce and
industry.  While straitened credit in one place may act to lower prices, it
may do this through distress liquidations, unemployment because production
cannot be financed, and various other working-capital effects.  (In a
"deficit" area, credit restriction may have perverse effects like hampering
local producers of import-competing goods.)  The long-term high interest
rates and restricted borrowing that you have seen in for example Argentina
have a terrible effect on businesses of all kinds.

International finance does have a rough center-periphery structure, with low
interest rates and relatively easy liquidity at the center, high rates and
endemic illiquidity at the periphery.  I have written more about this
elsewhere.  Not only would dollarization not change this, it is likely to
make it worse.

Best, Colin