Seriously, they have like ten papers on this topic from every angle. The one I
have doesn't do any of the history, just mentions Smith and Young once, and then
the bit on Arrow. I'd love to see the one you have. I never even heard of the
jl your talking about. Does your school subsc. to these?
I have drafts from several conference versions around somewhere. Growth and
technical change is of interest to them, as is the capital critiques, so when
the 'new' g.t. and new technological change stuff and 'endogenous' stuff started
getting thrown around they had to jump in. This one was summarizing is just by
Heinz.
By the way, Heinz knew Lowe well. He and Hagemann were young profs at Bremen
when he returned to Germany, and they were looking for their own forerunners so
to speak. The Italians had Sraffa, etc. It turned out that there was some
interesting relationships between Lowe's growth stuff and Sraffa's framework. So
they were working on Sraffa-Lowe-Keynes synthesis. Me the young brat got
intersted in Lowe but I started complaining that Lowe's methodological work
contained an implicit critique of the long period method. (actually I ended up
finding an explicit critique of it in Lowe from the thirties). This was a minor
annoyance to them, though later Heinz went strongly in the direction of Sraffa
and Harald cooled on the Sraffian line of things. In any case, what do you think
of Heinz and Neri's Theory of Production? Pretty impressive blurb from Samuelson
on it, dont you think?
-----Original Message-----
From: J. Barkley Rosser, Jr. [mailto:[EMAIL PROTECTED]]
Sent: Friday, February 16, 2001 3:45 PM
To: [EMAIL PROTECTED]
Subject: [PEN-L:8212] Re: RE: Re: Re: Re: new growth theory
Mat,
The one you have may be either an earlier version
or some variation. Mine starts with the Lowe quote also.
It took awhile for this paper to get published.
Among the earlier economists discussed at length
in the version I have are Smith, Ricardo, Torrens, Marx,
Charasoff, von Neumann, Kaldor, Young, Marshall, Wicksell,
Cassel, and Ramsey, all prior to Solow et al, much less
Arrow, Uzawa, Lucas, or Romer.
Barkley Rosser
-----Original Message-----
From: Forstater, Mathew <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED] <[EMAIL PROTECTED]>
Date: Friday, February 16, 2001 1:43 PM
Subject: [PEN-L:8197] RE: Re: Re: Re: new growth theory
>Jim says:
>
>>so what, in short, is the substance of their critique?
>
>I don't know that particular paper, but I have another one by Kurz.
>
>First, Kurz begins by quoting Adolph Lowe from a must-read 1954 article
called
>"The Classical Theory of Economic Growth" that of course I can't help but
>including here. Writing two years before Solow's famous article and thirty
years
>before Romer's dissertation, Lowe, with reference to some of the
then-recent
>contributions to mainstream growth theory writes that:
>
>"it is only fair to say that this modern notion of 'endogeneity' is but a
dim
>reflection of a much more ambitious method of analysis that dominated an
earlier
>epoch of theoretical economics. As a matter of fact, upon this issue of
>endogeneity versus exogeneity, rather than upon conflicting theories of
value,
>hinges the main difference between genuine classical theory and
post-Millian
>economic reasoning, including all versions of neoclassical analysis."
>
>Of course, Lowe considers Marx in this respect the zenith of 'classical
>political economy.'
>
>Kurz's article shows that Romer and 'new growth theory' are still but 'dim
>reflections', not adding anything substantively to our understanding of
>processes of growth and development.
>
>Kurz starts with Solow's model, relating it to new growth theory (NGT).
>Interestingly, Kurz shows that while one of the common views is that the
novelty
>of NGT is incorporation of increasing returns, IR is not an essential
>ingredient--if this assumption is abandoned, growth is no less 'endogenous'
in
>these models. He then shows that in NG models that start with intertemporal
>utility maximization, in order to get 'endogenous growth' the rate of
profit ( =
>mpk in competitive eq.) must exceed the rate of time preference. He
breezes
>through a whole series of variations, but I will skip ahead to a later
section
>that contain the main argument.
>
>In my earlier post where I argued that NGT is not new, I mentioned Smith,
Young,
>Kaldor, but I did not mention Arrow's 1962 contribution on learning by
doing.
>This is the starting point of Romer's 1986 argument.
>
>Arrow related the state variable 'level of technology' of a single firm to
>another state variable, the amount of capital accumulated in the economy as
a
>whole. In simplified form, output of firm i can be written as:
>
>Yi = A(K)F(Ki, Li)
>
>where the i's are subscripts, K aggregate stock of capital, Ki and Li
capital
>and labor employed in firm i. The increase in A is the unintended
by-product of
>the experience accumulated producing new capital goods. This learning by
doing
>is taken to be purely external to firms producing or using the capital
goods.
>(this strong assumption allows avoidance of the question who gets rewarded
for
>the increase in scale factor A!) There is the potential for endogenizing
the
>rate of growth in Arrow's model via a flexible saving rate, this was not
used
>because he assumed a fixed share of income saved and a fixed capital
>coefficient, the ratio which gives the growth rate.
>
>Romer's model differs from Arrows in essentially two respects: he adopts a
>conventional neoclassical production function and he assumes intertemporal
>utility maximization. Attention is on a single state variable: 'knowledhe'
or
>'information.' The main idea is that information (contained in inventions
or
>discoveries that have become innovations) is a non-rival good. However, it
is
>not totally non-excludable. Firms can exclude others from the info for some
time
>at least. This allows them to pocket temporary monopoly profits. The whole
>argument revolves around these two different aspects of the publicness of
the
>goods--non-rivalry and non-excludability.
>
>The basic idea of the 1986 romer paper is that there is a tradeoff between
>consumption today and knowledge that can be used to produce consumption
>tomorrow. To this effect he assumes a single research technology that
produces
>'knowledge' from foregone consumption [this is all so Irving Fisher!]
Romer
>conceives of 'knowledge' as a magnitude that does not depreciate and can be
>measured on a single scale as a continuous variable. (Kurz notes that these
are
>assumptions that are difficult to sustain: knowledge is not a variable that
can
>be measured on a single scale, it is in any case not cardinally measurable.
It
>is intrinsically heterogeneous. Different kinds of knowledge are
differently
>useful in production, new knowledge often renders previous knowledge
obsolete,
>etc.). In other words, it is like non-depreciating capital.
>
>Romer stipulates a research technology that is concave and homogeneous of
degree
>one. So,
>
>ki = G (Ii, ki) (where the k on the left hand saide should have a little
dot
>over it)
>
>where the i's are subscripts, ki is the current stock of private knowledge,
Ii
>is the amount of foregone consumption in research by firm i and ki (with
the dot
>over the k) is the induced increase in the firm's knowledge. The
production
>function of the consumption good relative to firm i is:
>
>Yi = F(ki, K, xi)
>
>where K is the accumulated stock of knowledge in the economy as a whole and
xi
>all inputs different than knowledge. These inputs are assumed to be given
and
>constant over time. "factors other than knowledge are in fixed supply"
writes
>Romer. Kurz says that this assumption is comprehensible for labor, since
one of
>Romer's arguments is that "population growth is not necessary for unbounded
>growth in per capita income. For simplicity it is left out." It is also
clear
>with respect to land of a single or different qualities, so that it is
something
>like Ricardo's idea of the "indestructible" power of land (for the sake of
not
>quibbling we'll leave the problems here aside). But Kurz writes that the
premiss
>does not apply to exhaustable or renewable resources or to capital goods.
The
>amounts of these cannot be kept constant by assumption. Strictly speaking,
Kurz
>writes, Romer's assumption implies that there are no exhaistable and
renewable
>resources or capital goods in the model. Put differently, there exists
only
>labor, Ricardian land, and 'knowledge.' That is, 'knowledge' is the only
>quasi-capital utilized in the production of the consumption good.
Spillovers
>from private research activities cause improvements in the public stock of
>knowledge K.
>
>Well, this continues with an investigation of various technicalities, for
>example whether the function is homogeneous of degree 1, < 1, or > 1 in Ki
and
>K, all of which give us different peculiarities, etc. For now, one of the
main
>things Kurz shows is that assuming, contrary to Romer, that the function is
>homogeneous to degree one in ki and K, the saving-investment relation
determines
>endogenously the growth rate and we don't need increasing returns to get
>endogenous growth, we get it even with constant returns.
>
>Now, a few of Kurz's other criticisms:
>
>1) since these models revolve around a few broad and rather obvious ideas
that
>have appeared again and again in the history of thought, the only thing
that is
>new is the formalizing of some of these ideas within a macroeconomic steady
>state framework.
>
>2) it can be doubted whether these formalizations add anything new to our
>understanding of growth processes
>
>3) no one could deny that STRUCTURAL aspects play an important role in
processes
>of growth and development. For example, changes in institutional framework,
>distributional shifts of output and employment between different sectors,
>different forms of technical and organizational change. NONE OF THESE PLAY
A
>ROLE IN THE NGT.
>
>4) as steady state models the only technical progress they can allow for is
the
>labor-AUGMENTING type. (the NGT and Solovian models share this important
>feature: they contemplate only a single form of overall technical change).
with
>regard to technical change the framework is a straight-jacket that
seriously
>limits the scope of analysis
>
>5) as regards preferences, in NGT consumption is taken to be the end of all
>economic activity. In some formulations, not even leisure is considered to
be a
>good that increases consumer utility. In these models, saving is carried
out, in
>a world of perfect foresight, solely for the purpose of changing the time
>pattern of the flow of consumption. EVEN IRVING FISHER was aware of
>non-consumption enjoyments. Of course, the classical economists and
especially
>Marx understood that the holding and accumulation of wealth for its own
sake was
>an important part of capitalism. When this is allowed for, certain familiar
>results of time preference theory that underlie most NGM cease to hold.
>Ironically, the strive to accumulate wealth (or some similar desire) would
lead
>directly to an endogenous explanation of growth!
>
>6) The process of economic growth is indissolubly intertwined with the
emergence
>of new methods of production and new goods, the coexistence of different
methods
>of production and the gradual disappearance of presently known methods and
>goods. Variety is an essential ingredient of growth. This fact is
acknowledged
>in some versions of NGT. Yet the way this is done reduces a world of
>heterogeneous goods, the variety of which may change, to a world with a
single
>good only. Indeed, to use Lancaster's term, all the different goods
produced
>and consumed represent different amounts of a single characteristic only.
If a
>chaning (increasing) variety of commodities due to product innovations is
sought
>to be covered, however imperfectly, by intertemporal utility maximization,
it
>would have to be assumed that all goods to be invented only in the future
are
>already allowed for in the current utulity function. Alternatively, there
must
>not be any genuinely new goods but only new products representing different
>amounts of a single characteristic.
>
>7) An aggregate production function is designed to represent complexy
>microrelations of production in simple terms. While the quest for
simplicity is
>laudable, in this case we know that the conditions required are excessively
>restrictive, never to be realized in reality, in the words of Franklin
fisher
>"far too stringent to be believable". Despite this NGT uses aggregate
>production functions.
>
>8) In NGM, capital, whether physical or human, is substantially upgraded as
a
>factor of production relatively to (simple) labor and to natural resources;
the
>latter is hardly ever taken into account, the former in some contributions
only.
>In one class of models there is even only a single factor of production
>'capital'. The entire capital controversy in ignored in NGT as if it never
>existed.
>
>9) Solow in his 1956 paper made it clear that his analysis was based on
>extremely simplifying assumptions. In the concluding section, entitled
>'Qualifications', Solow emphasized that "All the difficulties and
rigidities
>which go into modern Keynesian income analysis have been shunted aside. It
is
>not my contention that these problems don't exist, nor that they are of no
>significance in the long run. My purpose was to examne what might be
called the
>tight-rope view of economic growth..." The NGMs share this "tight-rope
view of
>economic growth" that is they set aside economic fluctuations and assume
that
>the economy follows a path characterized by FULL EMPLOYMENT OF ALL
PRODUCTIVE
>RESOURCES.
>
>10) The NGT has no proper analysis of investment behavior. It is a major
>shortcoming of both the Solovian and the NGMs that no serious attempt is
made to
>represent investment and to analyze the interlay of savings and investment.
To
>assume Say's Law is not good enough.
>
>There are a bunch more criticisms throughout, but these are the main
points.
>
>
