Re: Danby Seminar

[Bruce McFarling <ecbm@xxxxxxxxxxxxxxxxxxx>]

On Wed, 08 Oct 1997 11:09:14 -0700, Colin Danby <danbyc@xxxxxxx> write:

>...
>There's still a point here that I, at any rate, am not clear
>on and which maybe Bruce or someone else can help me unpack.
>It has to do with the nature of the time in which we think
>about a constraint and about liquidity.

>As I read the Thirlwallian argument, we have in essence a
>BoP flow constraint.  Imports plus capital acct balance
>and reserve use are all you have to pay for exports -- that
>kind of thing.  From this kind of approach we get the foreign
>savings gap in Chenery's 2-gap model and so forth.  In that
>sort of perspective we usually assume away "liquidity"
>problems properly speaking in the sense that if our period
>is one year and your exports are mainly in the 2nd half of
>the year, we assume that short term financing for imports is
>readily available in the first half (i.e. you can borrow
>on your 2nd-half imports in the first half.)

        But the Thirlwall law constraint is a long-period
constraint.  If a country or a group of country succeeds in
escaping a BoP constraint over a long period, some other
country or countries must have been left 'holding the bag',
and *are* facing demand constraints. That is, the two
alternative possibilities are that *everyone* is in BoP
balance over the long period, or that some are export demand
constrained because others are not.
        I have, in the past on pkt, argued for institutional
reforms that permit groups of demand constrained countries
reduce the impact of demand constraints (though this is not
how I originally articulated the policy), while paul has
been persistantly arguing for a more fundamental reform of
the international trade and payments regime to prevent
individual nations from straying far into the surplus side
away from BoP balance, and thereby permitting the system to
escape the limits due to the demand-constrains on individual
countries.
        I think the characterisation is that being export
demand constrained is *not* limited to less developed
countries, but also covers many high income industrialised
economies.  See the JPKE issue with the Thirlwall's law
minisymposium for more on this and to draw your own
conclusions.  But the Thirlwall's law argument is *meant*
to be a long period argument.  I reckon that *under* the
TL argument, de nile aint just a river in egypt: you
can facilitate growth with policies that increase the
income elasticity of your exports and/or reduce the income
elasticity of your imports, or you can pretend that the
long period problem can go away if you attract foreign capital
inflows and postpone the growth constraint for a while.
My questions were directed to thhe ways that financial
fragility might interfere with the first pair of strategies,
but I guess that financial fragility might also be exacerbated
by reliance on the policy of denial.

>When we move to a Davidsonian world things look different.
>Units have future payment commitments and cannot be sure
>enough of timing and extent of future income that they need
>to hold money and do various other things to avoid being
>caught short.  Liquidity is an ever-present problem but not
>a simple constraint as it was in the bookkeeping sense of
>the BoP.

        The Thirlwall's law argument is *within* Davidson's
world, in which the *lower* limit on a buffer stock and
the *upper* limit on a buffer stock are very much assymetric
boundaries.

>Take Indonesia, which has obligingly had a financial crisis
>this week in order to illustrate my arguments.  Indonesia
>has a fat trade surplus that is only going to grow fatter.
>Yet the currency is dropping like a stone, and liquidity is
>tight, in both $ and rupiah terms. Short-term interest rates
>in the Philippines are also absurdly high right now.

>Why aren't people rushing to buy rupiah at their cheaper $
>price?  Because they don't know how much damage a fragile
>financial structure is going to sustain, damage that
>will have a variety of real effects.

        Having a trade surplus does not necessarily
identify a country as not being export demand constrained
in the long period.  That is, Thirlwall's law is a BoP
argument, not simply a BoT argument.  If Indonesia roughly
has the trade surplus that it *has* to have to meet prior
commitments requiring capital outflows, then increased
difficulties in financing lead directly to greater
limitations due to export-demand constraints on growth.

>> However, the point is important *not* because any group
>> within the country can create liquidity within itself -- it can
>> only create a form a *local* liquidity. Rather, it is important
>> because different groups within the country face different obtsacles
>> (in both severity and type) in forming groups with foreign individuals
>> to create international liquidity.
>
>The liquidity problems I'm trying to zero in on are those
>affecting individual units, as Bruce notes.  At the moment I
>don't see those as flow constraints over long periods so much
>as weaknesses of financial structure that can produce nasty
>surprises in the very short run.

        I suspect interactions.  It's just trying to pin those
slippery beasts down that is hard.  So much easier to live in
a world of a few simple, universal, and eternal verities, that
leaves you forever to pin interactions down.

>So I see this as a structural constraint operating on
>individual firms, with further structural and political-economy
>consequences (like implicit state guarantees -- Daniel Kostzer
>made a related argument a couple of weeks back), rather than a
>flow constraint.  The 2 kinds of problems could interact in a
>number of ways; Bruce noted some short-term possibilities in
>his first post and I also noted that possibility that in the
>very long run a weak financial sector may be associated with
>a weak and capital-poor export sector.  But I don't have a
>neat theoretical way for relating the 2.

        I tend to think of this in terms of I-O relationships.
So the following is a tentative series of hypotheses that
might establish some sort of link.
        An export sector relying on high value-added use of
productive equipment and local techniques is either relying on
a local advantage in making use of imported technology, or relying
on a local advantage in making use of locally generated technology
(I'll stress again that I am using technology with its
social-science connotations, and not with its more narrow
engineering connotations).  But which is more likely to be
imitated by other *industries* locally, and which is more likely
to be imitated by other *countries* in the same industry?  That
is, which is more likely to serve as the basis for a wave of
innovation that is focused in the region / country in question?
If we hypothesize that a local advantage in using a locally
generated technology is more likely to generate a wave of
innovation with a local focus, then we have a link between the
locally generated productive equipment, techniques, and
organisation, the local environment for making use of such
locally generated technology, and effectiveness in terms of
increasing the value added of exports.
        The link between the financial sector and the export
performance is made if we hypothesize that foreign-based
firms have an advantage in the finance of imported technology,
and local firms have an advantage in the finance of locally
generated technology.  That seem plasuible to me, but it is
by no means certain.  But, under the hypothesis, a strong
locally based financial sector provides a part of the
environmental conditions required for the emergence of
a broad based local advantage based upon locally generated
technologies.  To be sure, if a strong relationship does
exist, it would far more likely be in the guise of a
necessary precondition than a sufficient precondition:
while a weak local financial system may be an impediment to
independent industrial technological development, a strong
local financial system is no guarantee of freedom from
dependent industrial technological development.
        And, to stress again, and as Hubert has noted,
the Thirlwall's Law argument is *not* a short period
argument, but a long period one.  So I did not intend
the possible interactions that I raised to be seen as
short period interactions, but as long period ones.




Virtually,

Bruce McFarling, Ourimbah, NSW
ecbm@xxxxxxxxxxxxxxxxxxx