RSS
Home
The missing economic model - part 1
Print
Email
RSS
Subscribe
A week ago Friday, my article was an observation that the tycoon style video games used a simplified economic model that created a misunderstanding about the economy. I noticed that after reading it Marty Carbone took the time to write a short question about that observation.
Are you suggesting there is a real-life model that can be used to explain our economy? Or are you saying our economy is too complex to understand? I am of the opinion that no economist really understands our system. If they did -- we would not be in this money and banking mess. If an economist understood our money and banking system -- he would certainly be trying to explain it -- in the same way that Carl Sagan explained everything –
Its actually one of the better questions about economic modeling I've seen, and the answer is rather complex. The simple answer is that Marty is correct. There is no real-life model that can be used to explain our economy. The complexity of the economy is bound up in the hundreds of decisions that millions of consumers each make every day. No mathematically model can capture every variable involved. However, the full answer must note that while the economy is extremely complex, the economists who do not try to model it often have accurate insights into what is happening.
The best way to answer this question is to break it into two separate answers. The first part is the issue of economic modeling; the second will be the understanding of the economy. Today, I'll cover the modeling part of the problem.
The difficulty with modeling the economy is just how interdependent so many parts of the economy are. Even when Adam Smith first decided to investigate "The Nature and Causes of the Wealth of Nations", he noticed that there was a great degree of connectivity between transactions even if the two transactions were conducted in different nations and weeks apart. Since then other economists have noticed that this connectedness generates changes throughout an economy even when only a small adjustment in price or technology occurs. A great example of how deep some of these connections run can be found in what happened with corn shortly after the US demanded that ethanol production from corn increase.
As one would expect more corn was being used to make ethanol, which reduced the availability of corn to be used as food or feed for livestock. But the changes rippled through the economy. As corn became scarcer and prices rose, more farmers began planning to plant corn. At the same time, food products using corn rose in price, sometimes dramatically even as ranchers complained about the lack of corn necessary to feed their livestock. The changes continued. As more fields were converted to corn farming, water usage changed, and other grain products became harder to find, even the farm machinery usage levels changed. Yet, its not possible to say with certainty how much of the changes were due to the demand of more ethanol, and how much was due to changes in prices of other commodities. It might be possible to ask one farmer to explain why he planted additional fields of corn in one year, but there are thousands of farmers, and each has their own reason behind the decisions they make.
If someone wanted to model the economy, they'd need to be able to attach a variable to every economic decision each person makes daily and find a way to isolate each variable, without making the isolation so complete as to lose the interdependency. And that is the problem with most models. Although the Keynesian equation is often written as Y=C+I+G+(X-M), most entry-level economics classes use two-dimensional graphs to illustrate the relations of variables in the economy. The basic Keynesian equation as I wrote it would require five dimensions, and if you break each of those variables down into their true components, you end up with a complex array that cannot be easily solved for a single variable. The mathematical models usually try to solve for a single variable by locking the other variables and assuming no changes. That method does not work.
But, that method is the standard one used by most government agencies including the CBO. The CBO has never met a tax increase they didn't predict would increase revenue and has always said a tax cut will resulted in decreased revenue. This occurs because they model the economy in a static mode and allow for no changes to economic behavior caused by the change in tax rates. It results in the office having a history of being consistently inaccurate when determining the true costs of any bill proposed in Congress. The inability of their models to predict possible changes in economic behavior means they fail to price in how consumers and businesses will react to the changes.
However, it's fair to say that these groups attempt their best when it comes to modeling the economy. The fact that the complexity of what they are modeling makes it impossible to capture every variable has resulted in economists using partial solutions. They are not perfect, but they allow some economists to make reasonable predictions based on what they can see.
Related Articles:
Add a Comment
