Thursday, 31 January 2013

A Supercomputer-Based Economic Model

It was less than thirty years ago that the American people elected a movie star to become President. His theory of  “supply side” economics, also called “trickle down economics” and “Voodoo economics” resulted in a short term spike in prosperity followed by a realization that we had run up a huge mountain debt for the country.  Meanwhile, in an attempt to keep up with the flood of U.S. government spending on defense, the Soviet Union pretty much went bankrupt and eventually went out of business, thus generating the new capitalist  nation we now call Russia.
Today, as our ship of state founders, having rammed an economic iceberg of frozen mortgages and credit, we are a divided nation – as least as far as economic theory goes. On one hand we have Republicans who, more or less, advocate more Voodoo economics by cutting taxes on businesses and the rich folk.  The theory is that if you give the rich folk a lot of money they will hire more people – instead of just banking the profits. On the other hand the Democrats are looking to FDRish, New Dealish, concepts including lots more government regulation of the financial sector, direct injections of money into the economy by construction projects, and other means.  Both Democrats and Republicans seem to like the idea of bailing out the banks some more, in one way or another. The question I have is this: isn’t this all a little Voodooish too?  I mean: where is the science?
In the United States we have quite a few supercomputers, massive conglomerations of processing units that can perform umpteen zillion computations per second.  They are great for predicting the behavior of nuclear reactions when there are lots of variables and lots of things happening at the same time.  They’re great for predicting the outcome of complex processes, like what will the weather be tomorrow, given that we have inputs, every second,  from hundreds of weather stations scattered all over the place.  So, if I might ask, why doesn’t our government create a vast working model of our economy in a supercomputer? Then the Democrats and Republicans could poke a variable here and there and see what comes out the end.  We could test John McCain’s economic theory and simulate the results with high reliability.  We could do the same with President Obama’s.  We could even test Rush Limbaugh’s – if he has one.  The problem is we don’t have a national or world economic model running in a supercomputer – at least I haven’t been able to find one, and I haven’t heard any politician saying to the other, “Hey, let’s put your idea into the national supercomputer and see what happens. Let’s see whose proposals will work the best, and then we’ll use that!”
It seems that Ben Bernanke does have a sort of economic computer model that works sort of OK.  But it isn’t a supercomputer model, capable of incorporating lots and lots of variables. There is a report that the Argonne National Laboratory is now beginning to take steps in the creation of an economic model that runs on their supercomputer, which is one of the fastest supercomputers in the world. This could be a major step forward in getting us away, for once and for all, from politicians making economic decisions based upon unproven ideology, constituent self-interest, or just Voodoo reasoning.  While we wait for the Argonne supercomputer programs to be written, perhaps we could ask the Democrats and the Republicans to give their suggestions into Ben Bernanke and have him put them in his program and see what it predicts – just for fun, you know?
It seems like such an obvious idea – using our national supercomputing capabilities that we have had for many, many years to model our economy. It makes one wonder why we didn’t do this twenty years ago. I suppose it’s a good thing we weren’t doing Voodoo nuclear engineering twenty years ago, but it’s long past time we got past Voodoo economics too,  and began studying economics as a multi-variable science, just like the weather.  With enough computing power, and the application of real science, we should be able to model, create, and maintain a stable, growing economy for a long time to come.  Too bad for the politicians, though, I wonder what they would have left to debate about then?

Rich McSheehy's Weblog

New Directions

PS. I added my link contact on Transfinancial Economics as comment to the original article above , and  it was accepted. Supercomputers play a vital role in advanced stage TFE

No comments:

Post a Comment