[From Rick Marken (2003.02.23.1130)]
Bjorn Simonsen (2003.02.23.09:00 CET)
I enjoy myself very much following the economic debate (Bill, Bill and
Rick). My leisure time is limited and I am still behind with all your
letters. If I become up-to-date with all the stuff you have presented I am
sure I'll participate more. This is interesting.
Perhaps the attached spreadsheet will make it a little easier to follow. I hope
many of you who are interested in the economics discussion will have Excel or a
spreadsheet program capable of running .xls files. Econ004RM.xls is my
transcription of Powers' ECON004b (which was written in Delpi, I believe) into
Excel.
When you open the spreadsheet you will be asked whether you want to enable the
macros. Say yes. Then you will see the main layout of the model in the
spreadsheet. The top three boxes are three sets of control systems: one set for
the plant manager (producer), another for the wage earning consumer and the third
(far right) for the capital income consumer. The three lower boxes are the three
main components of the economic "environment": plant, wage earning household and
capitol income household.
The control system variable names are pretty self explanatory. For instance, refRw
is the reference for variable Rw. The "environmental variables", like Rw, are less
self explanatory so they are commented (indicated by the red triangles in the
corner of cells). For example, when you hold the cursor over the red triangle in
the cell labeled Rw you see "Savings, $" so you know that Rw is the wage
consumer's savings in dollars.
You can enter your own parameters in the cells that are colored light green. For
example, you can enter a new value for refVm (which really should be called refVp
since it is the reference for Vp, the plant's inVentory of goods, but it's called
refVm in Bill's program so I'll keep it for now).
There are three buttons in the upper right. The "Run Model" button will run 2000
iterations of the model. Pressing this button always runs the model from whatever
state it is currently in (it starts the iterations with the level of cumulative
variables left from the last run, if there was one). The "Dynamic Graphs" button
shows a graph of selected variables, as in Bill's Econ004 model. After going to
the dynamic graphs you return to the "Basic Model" sheet by pressing the "Back to
Model" button next to the graphs. Finally, the "Reset Model" button resets the
variables in the model; in particular, it resets the cumulative variables to the
starting values shown in the "Accumulator Initialization" cells next to the
cumulative variables. If you don't press the "Reset Model" button, then each press
of the "Run Model" and "Dynamic Graphs" buttons runs the model starting at the
current state of all variables.
The parameters in the model are currently the same as those set in Bill's Econ004,
except for the reference for Vp (refVm) which is much lower than the value Bill
used. But I wanted to see if the model works OK when the plant manager has a
reference for inventory that is pretty low. I think refVp should probably be close
to 0, and the model works with this reference.
As far as I can tell, the model currently matches, in performance and details of
code, Bill's Econ004b version. I like the spreadsheet version of the model, not
only because the spreadsheet probably makes the model more accessible to more
people, but also because it makes it easier for me to see the relationships
between variables in the model and directly change the code that influences those
relationships. And Excel has a "trace" feature that puts arrows in showing
dependencies between cells. This makes it easy to see, for example, which
variables are affected by the outputs of the plant manager control systems
(outputVm, outputIm and outputRm).
Anyway, here it is. Enjoy. Experiment with new references and gains (the slowing
factors are currently irrelevant since all systems are proportional controllers at
the moment). Try different starting values for the accumulators. Most important,
try new relationships between economic environmental variables.
Dick Robertson,2003.02.22.1550CST]
So far, whatever the flaws in Rick's attempts to model leakage I was interested
in, and remembered well, the forecast with which he went out on a limb a couple
years back. Wherever it came from, it seems so far to have been rather
prescient.
Yes. I think Bush's persistent pursuit of tax cuts in the face of evidence that
they have no beneficial effect will test the leakage notion even further. As the
rich continue to get richer, leakage should continue to increase, resulting in
higher unemployment and inflation (both of which we are starting to see). I think
the dividend tax cut should really push it over the edge.
Best regards
Rick
Econ004RM.xls (95 Bytes)
···
--
Richard S. Marken
MindReadings.com
marken@mindreadings.com
310 474-0313