On Fri, 10 Feb 2012 14:16:57 -0600, Robert Grumbine
<
bo...@saltmine.radix.net> wrote:
>In article <
a579907e-e9a2-4b56...@j15g2000yqb.googlegroups.com>, TimR wrote:
>> On Feb 10, 8:03 am, "Steven L." <
sdlit...@earthlink.net> wrote:
>>
>>> Stagnant (or even negative) economic growth hits the poor the hardest,
>>> because it takes away their opportunities. Those who are already
>>> affluent have an easier time staying that way than those who are at the
>>> bottom of the ladder.
>>
>> Probably so, but it's circular logic. Operating within an economic
>> system designed to require growth, failure to grow obviously has
>> consequences.
>
> Steve should also clarify how that 'growth' is good for the lower
>80% of the economy -- who have seen their real incomes drop in
>recent decades, even though the total economy has 'grown'.
>
> I'm also still waiting for an advocate of 'growth' to define the
>term non-circularly. When there's economic 'growth', what is
>it that has grown?
When you put things this casually, it is hard to guess what you mean.
Economic growth is defined as an increase in output. Output is not
uniformly measurable, but it's at least somewhat measurable. The trick
for doing comparisons is to measure the same way consistently, but
there's nothing ambiguous or circular about saying something like:
Last year, I sold 1,000 widgets; this year I sold 1,100.
Whether or not you find a lot of economists working on the question of
economic growth is like the economy itself; it has its ups and downs.
Economists are more interested in the question of growth when there
hasn't been enough of it. For example, growth was a central issue in
works of classical political economy, including those of Adam Smith,
David Ricardo, Thomas Malthus, and Karl Marx. For most of that time,
economic growth was slow.
During what is known as 'the marginal revolution,' which saw the
development of general equilibrium approaches, economic growth was all
but ignored. During that period, economic growth in Western Europe and
the US was high and beginning with Marshall (1890), economic growth
was treated as exogenous-a very sharp break with the past. In
Marshall's fictional, steady state model, the economy grows at a
constant rate equal to the exogenous growth rate of the population,
which along with economic growth, was cooperatively high.
The basic structure of the general equilibrium approach was built on
thermodynamics. It should be immediately obvious from the term
'general equilibrium,' but if it's not, growth is not an issue in a
model of a system with minor variations around a state of rest.
Likewise (and despite claims to the contrary from some of the posters
here), growth has no part to play in the derivation of the capitalist
model, which is a straightforward extension of general equilibrium
with intertemporal borrowing and lending (consumption and saving at
the implied market rate of interest) via the Fisher Separation
Theorem. Even short-term disequilibrium economic models refer back to
general equilibrium. Ironically, in a perfect world the fictional
steady state model would be < drum roll > capitalism.
Exogenous Growth
Sometimes, people object to the exogeneity of growth in the
neoclassical model. If that is your issue, take heart. Keynes and some
others talked about economic growth, but they were, as Keynes said it,
all slaves to some defunct economist. That is, they talked about
growth from the perspective of general equilibrium. The Keynesian
model itself isn't growthy. It's a model of stabilization. However, it
does to some extent treat 'the long run,' so when John von Neumann
presented a general equilibrium model with endogenous growth in 1945
(A Model of General Economic Equilibrium. Review of Economic Studies,
13(1):1-9), it caused a buzz.
Endogenous Growth
While von Neumann's model was a simplistic application of linear
programming, it proved that in principle you could include growth in a
general equilibrium model. The derived growth rate in von Neumann's
model equals the interest rate. That's empirically wrong, but it was
interesting. Much more important, in the von Neumann model, economic
growth is endogenous, i.e., it is determined by what we do.
Sometimes, people object to the endogeneity of growth in modern
economic models. If that is your objection, it's hard not to share
discomfort about a subject that isn't going to be easily solved. The
endogeneity of economic growth makes it a contentious subject. Still,
endogenous growth that captures human activity has advantages and,
perhaps we can agree that it's better than postulating the raising of
a Paretian slab on the backs of eternally-breeding sky pixies.
On the other hand, if you were an economist we might not agree on
that, at all. Relatively few economists find the question of economic
growth interesting in itself. At best it gets in the way; at worst,
it's political, something most economists avoid for obvious reasons.
You will find most of them quietly doing empirical work or using a
theoretical approach in which growth is in the hands of the sky
pixies. On the plus side, comparative equilibrium statics is a
powerful tool for exploring what ought to happen when you do certain
things. Among other things, externalities and other forms of market
failure are revealed in this way. Most economists seem to find topics
other than growth much more interesting.
Intervening Years
Since Roy Harrod (1939) had earlier proposed a growthy version of the
Keynesian model, and because economic growth slowed with the mega
drought in the 1950s, growth became an issue again. Following papers
by Solow and Kaldor in the mid-1950s, endogenous growth was an
important part of economics until the late 1960s, when, following a
period of strong intellectual interest (and high growth), it moved
back to the fringe. That drought continued until the mid-to-late
1980s. Interest again fell off during the growthy Clinton years, but
has revived again recently.
Data-Driven Models
If you mean: "I'm waiting for an economist to unambiguously model
economic growth," don't hold your breath. Currently, there is no
coherent macroeconomic model of economic growth. There is no reliable
data-driven explanation for it, and there isn't likely to be just one.
Ever.
Data problems make up a minor part of that. As I said earlier, because
interest in economic growth has been episodic, so too is the data.
Most of the world has collected very little economic data. The data
that has been collected typically comes in difficult-to-use chunks,
often as short as a year or two spaced apart by a decade or more. That
makes cross-country and intertemporal comparisons technically
problematic and rife with conceptual difficulties. Even if you had the
most reliable data about past economic behavior that is currently
available, the most you could hope for is a broad idea about trends
and major differences.
On the other hand, data collection in the recent period is much
different from the past, and there is at least some data available for
most countries today. Not that it helps much. Back in 1991, World Bank
started a major research effort to understand why East Asia had grown
faster than all other regions in the last half of the 20th century.
Importantly, no common denominator for growth in East Asia versus
Europe and the US, which had the fastest growth in the prior two
centuries, was found. Here, the political trouble starts because, you
see, seemingly everyone has an opinion on economic growth regardless
of how little they actually know.
Clearing out (some of) the red herrings:
1) Population size, geographic size, economic size, or a combination
of the 3 does not change the rate of economic growth.
2) There are no caps on how high growth can be. Some macroeconomists
claim that countries in the developed world cannot grow faster than
about 3% per year. The usual figure is 6% for the less developed
world. Again, there are numerous counterexamples Ireland and
Luxembourg being very clear ones, but South Korea, Japan, Nauru,
Botswana, and Libya, are useful, too. Nauru and Botswana are
particularly important examples, because they lack nearly everything *
else* that is usually proposed to rationalize high growth in the
others (resources, liberal social policies, etc.). For ecological
reasons, there may be a cap on how much cumulative economic growth you
can have without a global collapse, but this is indeterminate.
3) Low efficiency neither necessarily decreases nor increases growth
Efficiency is important because resources are scarce, but it does not
in itself produce growth. For example, official growth in the USSR
averaged 15% per year over an extended period (the jointly produced
Western estimates were around 7% because so much of the output was of
quality too low to be usable). You see a similar pattern of low
efficiency, high growth, and rapid improvement in living conditions in
China and Vietnam under central planning. That does not imply that
inefficiency is good, but it may suggest that inefficiency may be
necessary to bypass entrenched interests.
In the first half of the 20th century in Indochina and along the
European-dominated Chinese coast, and the Manchu controlled interior
of China, regulation was light and efficiency was high, but growth was
very low, around 0.5% per year. In the second half of the 20th century
under central planning, growth in China and Vietnam averaged about 5%
per year and 3% per year, respectively. Peak rates have been higher in
both places. According to the IMF's data, China's average economic
growth in current US dollars has exceeded 12% per year.
On the other hand, 'socialist' central planning is hardly a panacea.
It fails more often than it succeeds. Still, efficiency is not a
particularly useful or unambiguous tool for predicting that growth
will occur. Likewise, there are long periods of high efficiency
combined with low growth in countries with various degrees of state
capitalism. There is no evidence of a relationship between them.
4) Low inflation is not a requirement for high economic growth. The
most obvious example of this comes from the Middle Ages, when both
growth and inflation were very low. Population growth appears to have
been low, too. By contrast, historically high inflation in the 20th
century is associated with high growth.
There are numerous recent counterexamples to the claim that low
inflation is essential to growth including the records of South Korea,
Saudi Arabia, Oman, and the USSR. On the other hand, there are also
problems with the claim that high inflation is good for growth, too.
Russia, Argentina, Chile, Yugoslavia, and Bolivia offer recent
examples. We have also seen high growth with low inflation in Germany,
Hong Kong, Singapore, Taiwan, and the US. The empirical correlation is
close to zero.
5) Next, you have a laundry list of claims about factors like culture,
ethnicity, language, religion, immigration, etc. Nothing there. When
growth is the highest in the world (or among the highest), pundits
claim that the culture/religion/etc. is superior. When growth slows
again, the label changes with it, i.e., the culture is no longer
considered superior, and past claims to that effect become the grist
of humor, viz, Texas a few years ago, or the antediluvian Old South,
the British Empire, Imperial China, the Conquistadors, and so forth.
Recently, we have heard claims concerning the superiority of
Euro-Socialism. Twelve years ago, unemployment in most of Western
Europe was sky high and the US was telling the Europeans to liberalize
their economies. There is no evidence that cultural factors are
relevant. That these claims are made by people with an axe to grind is
a red flag.
Similarly, historically speaking, heavy immigration is more commonly
associated with wage erosion, extreme poverty, and economic/social
collapse than it is with growth.
6) When someone claims that natural resources drive growth, ask them
to list Hong Kong's, Singapore's, or Luxembourg's, or why growth in
Angola and Nigeria is so low despite massive resources. Watch the
hand-waving begin, and the rush to explain using factors like culture,
ethnicity, religion, etc., all of which are empirically irrelevant
and/or indeterminate. The category of natural resources includes
things like geography. Poor Russia had no warm-water port, but then
neither do Switzerland, Luxembourg, or Austria. Throw in Botswana for
good measure.
I could continue in this vein for quite a while, but I don't think I
will. World Bank's data has been pretty thoroughly dredged and if any
economist could give you an unambiguous, answer based on the available
macroeconomic data, you'd have seen it already. Nevertheless, it's
good to have a list of things that don't determine economic growth,
even if saying so gets you screamed at by dedicated disciples of this
or that.
Vis-à-vis growth, for every macroeconomic factor you care to look at,
you will find examples where growth was high, medium, low, or
negative. So then, a list.
Keynesianism/Monetarism (these are versions of the same model): No
Geography: No
Population size: No
Population growth: No
Geographic size: No
Economic size: No
Culture: No
Religion: No
Ethnicity: No
Language: No
High/low efficiency: No/No
Low/high inflation: No/No
Immigration: No
Natural resources: No
Favorable geography: No
Education: No
Emancipation/education of women: No
Freedom/Democracy: No
Dictatorship: No
Income distribution: No
Health: No
Business performance: No
Just institutions/freedom from corruption: No
Infrastructure: No
Foreign investment: No
Banking performance: No
Stable politics: No
Science and technology: No
Mercantilism/neo-mercantilism: No
Pastoralism/agriculturalism/"Physiocratism": No
Liberalism: No
Import substitution: No
Comparative advantage: No
Foreign aid: No
High/Low interest rates: No/No
Low/High GINI: No/No
War: No
"Capitalism": No, albeit more often than Marxism/"Communism": Also, No
The list goes on. This, by the way, is exactly what you would expect
to find if economic growth is endogenous. It is, so if that's what
you mean by 'circular,' better get used to it. The people who cling to
one of these are not necessarily lying; however, they are uniformly
wrong most of the time. The correct answer could be "none of the
above," "one of the above," "any of the above," or, "some of the
above" depending on your circumstances and (unfortunately)
narrowly-scoped preferences. The only choice not on the list is "all
of the above." Some of the alternatives are mutually exclusive. None
of them are associated with rapid economic growth all, or even most,
of the time.
So, if your question is really just the common complaint that too few
people agree with your view, you might as well get used to that, too.
The 'other' answer is at least as likely to be right as yours, and the
people who believe it see their own motives and preferences as Good
and yours as Evil.
Again, the endogeneity of economic growth just means that it is
determined by what we do. In other words, it's not that we can't
control it, but that we have to, if we can. How to do that is a
rational question to ask and it may be worthwhile to answer it in
broad-brush strokes.
The answer itself is deceptively simple. Economists who insist that
growth is endogenous, and those who treat it as exogenous, are both
right in a sense. If you force growth to be exogenous, or sufficiently
exogenous, it is at least somewhat closer to straightforward to
control and direct it. Perversely, actually doing that pulls economic
growth out of economics and lands us squarely in politics, which is,
as I suggested earlier, something most economists avoid like the
plague. Mileage varies on how well political policies have worked out.
Sufficiently Exogenous Economic Growth
There are different possible uses for the term "sufficiently
exogenous," but as far as I know its first use by economists traces to
a 1975 paper by Richard Sylla, The American Capital Market, 1846-1914:
A Study of the Effects of Public Policy on Economic Development. It is
arguable that while the economy adjusts rapidly to an exogenous shock,
the adjustment might not be quite instantaneous, and that might give
you an opportunity to peak behind the veil.
The trouble with using 'sufficiently exogenous' as a term of art is
that it is undefined. The fact that one can get a room full of peers
to accept that it might be true doesn't make it so, and roiled waters
are murky anyway. It needs a bit of rigor.
Merriam Webster defines economics as "a social science concerned
chiefly with description and analysis of the production, distribution,
and consumption of goods and services." In broader terms, economics
describes the menu of human survival strategies.
Here we are in luck; biologists have been very interested in survival
strategies, and we can start there. The metaphor of the fitness
frontier is particularly useful. Although 'fitness' in these models is
given (exogenous), and some of the randomly mutating elements in them
are arbitrarily fit, that does not imply that thinking in terms of
fitness landscapes has nothing to do with biology, although,
ironically perhaps, that may have more to do with what is wrong with
them.
In the recent thread on fitness landscapes, there were two main
objections: 1) real landscapes are discontinuous and irregular, so
there's no defined slope for a hill climber to climb; and, 2) species
don't find fitness peaks, anyway. Both objections are likely true.
Neither is very important. Modern search algorithms use heuristics
that incorporate random mutation, so they aren't hill climbers,
anyway. The real problem is that stochastic optimizers don't find
optima efficiently in large, dynamic state spaces. The larger the
state space, the more complex the strategy tends to become, so instead
of optimal or near-optimal solutions, you get 'good' performance for a
flavor of good that you find acceptable.
Consequently, suggesting that simple organisms find optima is a tough
sell. Analogously, it is more reasonable to claim that the organisms
we observe found acceptable outcomes. Those that did not find
acceptable outcomes are not usually observed. While some of the
observed outcomes are better than others, 'peak' is probably not an
apposite word choice vis-à-vis biology.
Nevertheless, from thinking about possible solutions to a large-space,
dynamic fitness landscape, a number of observed strategies immediately
spring to mind including:
1) Wide dispersal of seeds and spores
2) Random search (ants)
3) Variants of optimal foraging (wasps, rodents, birds, and others)
4) Experimental probability (birds, probably)
5) Restriction/control the state space (humans, perhaps badly, and to
a lesser extent, others)
Although in talking about economic growth we are mainly interested in
(5), all of these items of interest should be related in the fitness
space. And, of course, on the empirical evidence, they are. They all
exhibit proportional hazard.
[1]
http://www.ncbi.nlm.nih.gov/pubmed/20737851
[2]
http://www.jstor.org/discover/10.2307/4376?uid=3739704&uid=2129&uid=2&uid=70&uid=4&uid=3739256&sid=21100705797501
[3]
http://espace.library.uq.edu.au/eserv/UQ:10184/poss4.pdf
[4]
http://www.ncbi.nlm.nih.gov/pubmed/20175592
http://www.individual.utoronto.ca/jayb/Publications_files/Biernaskie_etal_09.pdf
[5]
http://www.huduser.org/portal/publications/pdf/jic_sl_gk_rnr_rep_0903.pdf
In terms of the fitness landscape, if we had very powerful
instrumental variables, we might be able to extract a small part of a
superfit cohort of possibilities and some points on the fitness
landscape for a single state space. While that might yield some
interesting results, the effort would be very time consuming and
costly. Alternatively, we can treat fitness as a parameter, generate a
landscape, use it to model evolution, and get useful results at low
cost. The latter makes rational sense and, while it does not have
'everything' to do with biology, it has a great deal of 'something' to
do with it.
On one hand, it is not a surprise to find that organisms in complex,
dynamic state spaces follow strategies dominated by proportional
hazard. Whenever the state space is too complex to control, that
includes the human strategy of restricting it. For example, we might
expect major cities to exhibit proportional, spatial hazards because
topography is largely beyond our control, and can't be meaningfully
planned for. Likewise, the city gets to geographical barriers when it
gets there, if at all. Dispersal of seeds and spores, random
topographical search, optimal foraging, or finding things by
experimental probability will exhibit proportional spatial/temporal
hazard for similar reasons. Laissez-faire capitalism fits very nicely
with these other survival strategies.
One might go slightly farther here and guess that survival behavior is
always predominantly characterized by proportional hazard.
Mathematically, that is because dynamic change in the state space
makes optimization too difficult and/or inefficient. Some form of
heuristic will dominate the landscape, and a quasi-optimal process for
forming heuristics without knowing the entire hazard function makes
sense.
The Design Problem
We can make life more interesting in other ways, too. Hazards come in
a variety of flavors. I've mentioned spatial and temporal varieties,
but hazards can also be proportional or accelerated, as well. The main
advantage of proportional hazard lies in not needing to know the
hazard function. The main disadvantages are that they are difficult to
specify and that they perform poorly when misspecified, or confounded
by nuisance covariates. By contrast, the main benefit to accelerated
failure time models is that under certain circumstances, some model
parameters can be robust to omission of covariates, i.e., we can get
good answers about some things without knowing or being able to
control all possible causes.
In a complex setting, not needing to know the underlying hazard
function is a huge advantage; one can use proportional hazards to
approximate heuristics, but not more than that. The main problem with
the analysis of proportional hazards is that (often) nuisance
parameters get in the way.
For the ants in Haccou and Hemerik (1985), hazard falls (odds of
finding a tasty cinnabar moth larvae rises) when the larvae are
migrating between plants, i.e. on the ground, and with greater search
distance from the nest.
|x
| x
| x
H| x
A| x
Z| x
A| x
A| x
R| x
D| x
|
|
|________________________________
Distance From Nest
Since my current apartment is infested with (smelly) sugar ants, I did
some simple experiments, none of which yielded shocking results. I
scored an old cutting board into a 20x20 grid centered on the spot I
had observed them emerge and put a bottle cap on each square. I placed
a drop of honey in a random bottle cap in the center section of the
board every day for 10 days. I observed that ants visited most squares
every day, but most frequently in the squares that had had honey
before and of course, most heavily in the square that had honey on the
particular day.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
|/|/|/|/|/|/|/|/|/|x|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|x|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|x|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|x|x|x|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|x|x|x|1|x|x|x|/|/|/|/|/|/|/|
|/|/|/|/|x|x|x|3|x|x|x|x|x|x|x|/|/|/|/|/|
|/|/|x|x|x|x|x|x|x|x|x|x|x|2|x|x|x|/|/|/|
|x|x|x|5|x|x|x|x|x|6|x|x|x|x|x|7|x|x|x|/|
|/|/|x|x|x|x|9|0|x|x|x|x|8|x|x|x|x|/|/|/|
|/|/|/|/|x|x|x|x|x|x|x|x|x|x|x|/|/|/|/|/|
|/|/|/|/|/|/|x|x|x|4|x|x|x|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|x|x|x|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
x = 10 or more visits
/ = 0-9 visits
Not that this is exciting; it just means that you can get the ants to
follow a pattern.
These particular ants are also convenient subjects because they don't
change their foraging pattern very much. What might you predict my
ants might do if I put food in a new cell (marked 'f') on successive
days?
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
|/|/|/|/|/|/|/|/|/|x|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|x|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|x|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|x|x|x|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|x|x|x|1|x|x|x|/|/|/|/|/|/|/|
|/|/|/|/|x|x|x|3|x|x|x|x|x|x|x|/|/|/|/|/|
|/|/|x|x|x|x|x|x|x|x|x|x|x|2|x|x|x|/|/|/|
|x|x|x|5|x|x|x|x|x|6|x|x|x|x|x|7|x|x|x|/|
|/|/|x|x|x|x|9|0|x|x|x|x|8|x|x|x|x|/|/|/|
|/|/|/|/|x|x|x|x|x|x|x|x|x|x|x|/|/|/|/|/|
|/|/|/|/|/|/|x|x|x|4|x|x|x|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|x|x|x|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|f|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
x = 10 or more visits
/ = 0-9 visits
If you guessed that the foraging pattern would stay pretty constant,
but bulge in the direction of the new food site, you'd be right.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
|/|/|/|/|/|/|/|/|/|x|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|x|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|x|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|x|x|x|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|x|x|x|1|x|x|x|/|/|/|/|/|/|/|
|/|/|/|/|x|x|x|3|x|x|x|x|x|x|x|/|/|/|/|/|
|/|/|x|x|x|x|x|x|x|x|x|x|x|2|x|x|x|/|/|/|
|x|x|x|5|x|x|x|x|x|6|x|x|x|x|x|7|x|x|x|/|
|/|/|x|x|x|x|9|0|x|x|x|x|8|x|x|x|x|/|/|/|
|/|/|/|/|x|x|x|x|x|x|x|x|x|x|x|/|/|/|/|/|
|/|/|/|/|/|/|x|x|x|4|x|x|x|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|x|x|x|x|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|x|x|x|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|x|x|x|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|x|x|x|x|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|x|x|x|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|x|x|x|x|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|x|x|x|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|x|/|/|/|/|/|/|/|/|/|/|
|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|/|
x = 10 or more visits
/ = 0-9 visits
Presumably, the ants won't search any further than they have to, and
(evidently) it's not quite true that they don't change their foraging
pattern. Actually, if you wait a few days, erase their pheromone
trails, and make the food visible, they do less random searching. The
first experiment looks like a proportional hazard. The second looks
like accelerated failure. In the latter, most of the parameters have
been eliminated; the food was reliably in the same place.
If you did not previously read my post on urban economics, you can
find it here:
http://groups.google.com/group/talk.origins/browse_thread/thread/95a3068f18f261a2/aae603026d18b35d?lnk=gst&q=beck+and+city#aae603026d18b35d
This is the standard economic model of the city with its roughly
circular shape of concentric rings:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
|_|_|_|_|_|_|_|_|_|x|_|_|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|_|x|x|x|_|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|x|x|x|x|x|_|_|_|_|_|_|_|
|_|_|_|_|_|_|x|x|x|x|x|x|x|_|_|_|_|_|_|
|_|_|_|_|_|_|x|x|x|x|x|x|x|_|_|_|_|_|_|
|_|_|_|_|_|_|_|x|x|x|x|x|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|_|x|x|x|_|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|_|_|x|_|_|_|_|_|_|_|_|_|
And, this is an idealized city with N-S and E-W highways:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
|_|_|_|_|_|_|_|_|_|x|_|_|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|_|_|x|_|_|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|_|x|x|x|_|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|_|x|x|x|_|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|_|x|x|x|_|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|x|x|x|x|x|_|_|_|_|_|_|_|
|_|_|_|_|x|x|x|x|x|x|x|x|x|x|_|_|_|_|_|
|_|x|x|x|x|x|x|x|x|x|x|x|x|x|x|x|x|_|_|
|_|_|_|_|x|x|x|x|x|x|x|x|x|x|_|_|_|_|_|
|_|_|_|_|_|_|_|x|x|x|x|x|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|_|x|x|x|_|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|_|x|x|x|_|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|_|_|x|_|_|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|_|_|x|_|_|_|_|_|_|_|_|_|
|_|_|_|_|_|_|_|_|_|x|_|_|_|_|_|_|_|_|_|
The rule in real estate is: If you build it, they will come. One of
the main reasons for building highways is that they provide for
accelerated failure, i.e., they are corridors of relatively low
uncertainty. Access stimulates housing, shopping malls, and so on. The
overwhelming majority of the change in US population density since the
1950s follows the interstate highway system. At a glance, there is not
much difference between the humans and the ants, other than that we
have some idea where to put the sugar. On the downside, this also
works against us in some respects. The ultimate cause of urban sprawl
is the uncontrolled flow of subsidies from the urban centers to the
suburbs. In other words, it is analogous to continually laying sugar
on the expanding borders of ant-world.
The obvious temptation was and is to stimulate growth by building more
highways, but that will probably only work extremely well once.
Estimates of the social welfare return on investment of the interstate
highway system vary from 35%-60% per year in the early period, falling
to about 10% per year now. Most of the low hanging fruit has already
been plucked, though there are still some large cities, without direct
connections. Estimates run as high as 35% per year in the initial
period after improvement of the system, but that is a short-term fix.
Unfortunately, there are diminishing returns so that at the margin,
sprawl is offset by decay in the urban centers. Urban decay demands
more subsidies and, because subsidies flow outward from the urban
centers, that tends to make the problem worse. Nevertheless, managed
economic growth is largely a matter of accelerated failure. That is,
you decide what you want and decrease uncertainty to make that choice
easier for people to make. Parenthetically, that doesn't require
incentives; it can mean laying out some sugar or making growth flow
from the barrel of a gun.
This, by the way, is also the central secret of the success of
capitalism: The capitalist may be wrong, but he knows what he wants,
and he won't look further than he has to. Where he invests his savings
and/or those of others, there is, for a time, a zone of accelerated
failure, and even if long-term success is unlikely, no one has a
stronger incentive to succeed; the 'capital' in capitalism is, after
all, the investable savings of ordinary people. It is also why 'crony
capitalism' (often) works; it combines the savings of ordinary people
with subsidies from the state.
Likewise, it explains both the success and failure of central
planning. Central planning is inefficient, but you can do pretty well
as long as you can figure out where to sweeten the pot and/or point
the gun. [Not that there has been no use of guns under capitalism. For
obvious reasons savers are reluctant to give up both current *and *
future consumption.]
The weakness of central planning is that the state space changes
dynamically and, not being much smarter than ants, the planners
(mostly scientists in the case of the USSR, China, and Vietnam, by the
way), soon run into trouble. Widespread collapse usually follows, and
when it does, the political tug-of-war intensifies between groups who
think that you should do more of the same thing, do nothing, or do
something else.
Historically, the track record of planners isn't great-with good
reason. Nuisance parameters and dynamic states are a bitch, and you
can't usually ignore them. One of my more recent favorites is the
politically correct idea that growth and prosperity follow educating
women. Since economic depressions are symptoms of wage erosion, which
in turn stems from a labor surplus, whether domestic or foreign,
increasing the labor supply makes the problem worse, not better. Not
surprisingly, the empirical evidence says that pursuit of admirable
social goals follows rather than precedes strong economic growth. In
case you're wondering, that includes environmental issues qua the
environment itself. Slow economic growth is death to consensus on
environmental problems, and growth has been slow for a decade despite
massive domestic spending and two wars.
…Which gets us back to why most economists ignore economic growth most
of the time. Deciding what you want and making it possible is the
milieu of politics and politicians. Absent a dictator, that means
formulating a compelling vision and building consensus. Even that
doesn't necessarily work; consensus can be just as wrong as a
dictator.
That in turn, should tell you at least part of why the economic
recovery has been slow. The industrialized economies are heavily
exposed to a global labor surplus. Our sitting President is fairly
weak and has articulated no vision. Likewise, the opposition has
gridlocked policy wherever they can. Just as rapid growth occurs with
decreased uncertainty, increased uncertainty stops it.
In generic terms then, one way to look at the design problem boils
down to a reduction in the number of parameters necessary to achieve a
particular outcome with acceptably high probability. This is the
design matrix from classical statistics, and the test for parameter
reduction is familiarly attributed to Wald. If we further define
"design" as the conscious realization that reducing the parameter
space leads to a more efficient result than random chance, we find
ourselves on Bayesian turf and the question becomes one of reducing
the parameter space from what-to-what.
In evolving systems, the reduction seems to be from proportional
hazard to accelerated failure. Here, of course, we hit a snag because
yesterday's accelerated failure is an aspect of today's proportional
hazard. In the case of the city, the addition of new highways gives
you a temporary period of accelerated failure.
Highway
Accelerated
failure
^
/||\
/ || \
/ || \
/ || \
/ || \
/ || \
/ || \
/ City \ Highway
<==========================> Accelerated
\ Proportional / failure
\ hazard /
\ || /
\ || /
\ || /
\ || /
\ || /
\||/
However, since the rule is 'if you build it, they will come'
proportional hazard gives way to proportional hazard, thus:
Niche k+2
Proportional hazard
City (t2)
\
\ Niche k+1
Accelerated \ Proportional hazard
failure \ City (t1)
(E-W highway) \ /
\/
/
/
/ Accelerated
/ failure
/ (e.g. N-S highway)
/
Niche k
Proportional hazard
City (t0)
In biology, you might draw something like:
Vacant niche k+2
Proportional hazard
Stasis
\
\ Vacant niche k+1
Accelerated \ Proportional hazard
hazard \ Stasis
Speciation \ /
\/
/
/
/ Accelerated
/ hazard
/ Speciation
/
Vacant niche k
Proportional hazard
Stasis
…which looks about the same. As it should. Economics is a description
of the human survival strategy and is properly part of biology; it's
not obvious that the human survival strategy should look different
from any other, even if the proportional hazard we face is a hybrid of
natural proportional hazards and designed, accelerated failures. Other
organisms find accelerated failure too, of course, if not by design,
then by random mutation and natural selection.
As an aside, the problem of design is a conflation of an perceived
decrease in randomness with a reduction in parameters-a sort of
parametric shortcut. The former does not imply the latter. One of my
favorite examples is a photograph of a naturally eroded piece of
sandstone that looks like the carved face of a Native American
tribesman. Since when asked, some people guess that it's natural and
others guess that it was designed, pundits claim that it proves that
you can't detect design.
Incredibly, they make this claim having detected their designed
computer, detected and connected to the designed Internet, and having
found a counterpart in the ether who has just done the same thing and
is more or less simultaneously examining a digitized photographic
image. The notion that design is undetectable borders on silly; almost
all people do it every day. We program computers to do it. These days
we program them to infer aesthetics, too.
Unfortunately, the brain says 'it looks nonrandom, so it's probably
designed,' rather than 'it would require controlling an enormous
number of parameters to make this, so it's probably natural.' However,
that the brain over-detects design and/or opportunities for design is
a different question from whether it detects design at all. My senses
don't tell me directly that the earth orbits the sun, either. That's
not to say that we couldn't design a sand blaster and, at substantial
cost in time and treasure, etch faces into sandstone. Ironically, the
hallmark of good garden design in China and Japan is making something
artificial appear natural, so we might even be able to sell them.
Here, I have to admit some ambiguity, because if we ever saw a second
etched sandstone face just like the first, we would very strongly
suspect manufacturing and some associated design, even if the main
shortcut is only production time. For the same reason, I don't doubt
that we could, at great cost in time and treasure, painstakingly
reconstruct a person's DNA (or a suitable template) one base pair at a
time. On the other hand, since we can't seem to make our trains run on
time, our occasional errors would probably make it excessively hairy
or have Wilkin's knees. The creationist claim that plants and animals
are too complex to have evolved is then, conflated and backward. If
there is a great Designer-in-the-Sky, the proper question to ask from
one designer to another is: What was your shortcut?
Evolving Economies
Returning to economic growth, if by 'circular' you mean 'moves from
one SNAFU to another,' that's true, and I can't at the moment imagine
a world in which it would not be so. Might as well get used to that,
too. Economics is forking in this direction.
One of the main problems in macroeconomics stems from the limitations
of the approach economists have tried to follow, i.e., merge the idea
of marginal analysis/general equilibrium with growth to create dynamic
economic systems by intertemporally optimizing changes in the economy
along optimal equilibrium trajectories. Models of this sort are
complex systems of differential equations, but the approach is still
normative; economic development is fully predetermined. They are
models of how the economy ought to behave rather than how it actually
does behave. They're not entirely new, but an extension of the Fisher
Separation Theorem.
One possible alternative to a purely rational, goal-driven model of
the economy is to treat the elements of the economy as coevolving.
These developments in economics mirror changes in another discipline
you seem to dislike, philosophy, specifically the philosophy of
design.
>And what relationship does that have to how well I live?
Begin by reading the following two passages.
In 1893, Supreme Court Justice David J. Brewer said: "It is the
unvarying law that the wealth of the community will he in the hands of
the few. . . . The great majority of men are unwilling to endure that
long self-denial and saving which makes accumulations possible ... and
hence it always has been, and until human nature is remodeled always
will be true, that the wealth of a nation is in the hands of a few,
while the many subsist upon the proceeds of their daily toil."
[Quoted in Howard Zinn, A People's History of the United States,
Harper Perennial Classics Edition, 2005, p261]
In 1879, Henry George (Progress and Poverty) said: "It is true that
wealth has been greatly increased, and that the average of comfort,
leisure and refinement has been raised; hut these gains are not
general. In them the lowest class do not share.... This association of
poverty with progress is the great enigma of our times.... There is a
vague but general feeling of disappointment; an increased bitterness
among the working classes; a widespread feeling of unrest and brooding
revolution.... The civilized world is trembling on the verge of a
great movement. Either it must be a leap upward, which will open the
way to advances yet undreamed of, or it must he a plunge downward
which will carry us back toward barbarism...."
[Quoted in Howard Zinn, A People's History of the United States,
Harper Perennial Classics Edition, 2005, p264]
Both of these were at least partly realized, albeit in different ways.
The first was true at the time and is still true today, mainly because
it captures human nature very accurately. The market between savers,
who delay gratification more, and spenders who are impatient to
consume more now, determines the prevailing rate of interest. Most
people save very little. Some people save a lot, often for multiple
generations.
The latter prediction also came true, but only in part, chiefly
because we now force people to save more. Most of that saving becomes
available in retirement, so for most people economic growth
translates, not so much into how well they live now, but whether or
not they can live at all later.
That should not be taken to imply that I think you're getting a good
deal. The Social Security system has never been actuarially sound.
Forced saving in the form of payroll taxes have risen dramatically to
cover both the cost of higher benefits and longer life expectancy. Had
the social security system invested in equity from the beginning, you
would be a lot better off now; t-bills don't produce a very good
return. At the time the system was created, equity ownership was
considered far too speculative for pension funds. At the time most
pension funds were invested in bonds, particularly those of the
railroads, which were considered very low risk securities, but that is
another story.
Nevertheless, your question boils down to a demand that you have your
cake and eat it, too. That isn't one of the possibilities on the menu.
Another way to look at it is that the majority of people have rejected
both collective and individual ownership of the means of production as
too risky. As a result, they have both less risk and less return. With
the possible exception of some tax-deferred hedge fund assets
(debatable), getting less risk and more return is also not one of the
possibilities on the menu.
As for this business:
> Steve should also clarify how that 'growth' is good for the lower
>80% of the economy -- who have seen their real incomes drop in
>recent decades, even though the total economy has 'grown'.
As of the 2010 Census, the compound growth rate in mean real incomes
for all 5 income quintiles in the US has risen since 1967. People in
the lowest quintile had a rough time in the last decade. That makes it
possible to cherry pick the data. For example, the biggest losers over
2008 through 2010 were the top 5%. Their mean real incomes were down
almost 11%. Not that I'd feel sorry for them; their geometric average
growth rate in real income since 1967 was also highest, at 1.4% per
year. Both are a consequence of their willingness and ability to bear
equity risk.
US Real Mean Household Income Growth per year by Quintile, Constant
2010 US$, over 1967 through 2010.
Q1 Q2 Q3 Q4 Q5 T5%
Mean(%/yr) 0.5 0.3 0.5 0.8 1.2 1.5
StDev(%/yr) 3.0 2.3 2.1 2.0 2.7 4.3
Compound(g) 0.4 0.3 0.5 0.8 1.2 1.4
Source: US Census Bureau
The data is freely available here:
http://www.census.gov/hhes/www/income/data/historical/household/
In order to compensate for a variety of leakages, growth has to be
fast and sustained to substantially improve the lot of the poor
because they do not save, but depend on rising real income from wages.
Empirically, slow growth 1%-2% per year is associated with declining
standards of living. In the developed world 'high' means more than 3%
per year. Historically, countries in the middle and lowest tiers have
required sustained growth rates above 5% per year and 7% per year,
respectively, to increase living standards for the poor.
The trouble with thinking that the solution to poverty is to ramp up
economic growth is that population growth tends to offset rising
income, so that per capita disposable income can actually decline as
overall income rises. Still, it is better than the alternative. The
effects of slow economic growth combined with population growth can be
pronounced, even in the developed world. For example, in Western
Europe average economic growth of about 2.5% per year saw unemployment
rise from 4% of the workforce to over 10% of the workforce from 1980
to 2000.
One solution to the problem of wage erosion via
population-growth-related leakages has been protectionism. It's a
mixed bag. It distorts your domestic economy, drives up the prices of
goods and services, undercuts saving, and beggars your neighbor, while
doing little or nothing to solve the underlying population growth
problem. Some of the Europeans, especially Germany, have been
protecting their markets against Chinese goods lately; however, most
of the world has protectionist for more than a decade. Like China,
they are manipulating their currencies, too. That has left the US
bearing the brunt of the cost of China's development. That's not the
worst thing; Europe is probably too precarious to do it. The situation
remains dangerous.
As I have said elsewhere, if you are going to protect your markets and
pay higher prices anyway, you might as well get something in exchange.
Just as urban sprawl is a consequence of uncontrolled subsidies,
overpopulation is a consequence of using children as income producing
capital assets. If you want to avoid wage erosion, you will eventually
insist that your trading partners:
1) pay a minimum wage
2) ban child labor, and
3) establish pensions for the elderly
Failure to address the actual problem will not make things better, and
will probably make things worse