Reading Assignment #3, "What Social Science Does- and Doesn't- Know"

A.

Truthfully, there was not too much I found interesting with this article, but there were two specific parts from which I did take something.

First, I found it interesting how the article argued that there is no sure-fire way to test social science/economic bills/programs because unlike natural sciences, you can't control everything that the bill/program is effecting. This leads to not being able to find definitive results when trying to test whether or not a social science program for improve will work successfully or not- at least without letting a program run its course for a significant amount of time.

I found this to be interesting because in today's society, so many people act as if they know what is the best way to improve our economy. Politicians trying to win office will promote a certain stimulus plan that they are confident will improve our economy.

Regardless of what George W. Bush may have claimed, he had no way of
knowing how successful- or unsuccessful- one of his bills/economic refo-
rms was going to be until he allowed it significant time to run its course.

But this article has made me realize that while there may be some reason to believe such politicians, the bottom line is that it is ultimately impossible for anybody to predict how successful (or unsuccessful) a bill/program will be, and therefore, it is important to take those claims with a grain of salt.

One other aspect of the article I found interesting was how the article suggested that over time/through social science experimentation, it was determined that changing one's environment is almost always more successful at fixing a social/economic problem than trying to changing a person.

For example, the article discusses a "social science experiment" about criminals and trying to reduce the crime rate in America. Following the "experiment", it was found that trying to counsel the criminals and changing their outlook on life really didn't change anything and crimes continued to be committed. But when they changed their environment- perhaps by moving them to a different, safer, city, crime was indeed reduced.

In another example, the article mentions how an economic program that creates more jobs was more effective in improve the economy than was a program that helped to build peoples' skills. Changing peoples' skills did not improve the problem, but by changing the environment- and adding more jobs- there was an improvement in a problem.

B.

1. Why is that people continue to be influenced by- and believe- politicians/governmental officials who claim they know how a certain economic program will work? It is clear from this article that in the majority of instances, it is imperative to allow a program to play out before evaluating how successful it is, so why is it that people continue to listen to these types of claims?
2. Since we can't experiment and find a definitive answer about whether or not a bill/program will be successful, on what do people base their beliefs regarding whether or not they think a certain program will improve a problem?
3. The article claims it is very difficult to accurately predict how successful a program will be. Is this really the case? Do you think that one can actually predict the success of a program after doing sufficient research and experimentation and if so, what is the best way to go about performing social science/economic experimentation?

C.

Overall, this article tried to summarize one main point: It is very difficult- or perhaps even impossible- to accurately test social science/economic reform with experimentation, as is does in the natural sciences. Because of this, it is imperative to give new bills/programs time to run its course before claiming that a program is either successful or unsuccessful.

Experimentation works in natural sciences because of how a "control" can be attained. If one does an experiment on let's say a plant, one can control the plants environment completely. If a person wants to see how a certain type of nutrient effect the growth of a certain plant, the person can pretty much get accurate results because they can do everything in their power to make sure that each plant is living in the same exact environment and receiving all the exact same treatments as another plant, except for the nutrient it is receiving. Therefore, from the experiment, one will learn a definitive answer about how successful a certain nutrient is at nourishing a plant.

In economic/social science bills/programs meant for improvement, you simply cannot do this. Let's say the government is trying to improve education in America, and it passes a certain program it thinks will help education in America imrpove. There is no definitive way to know whether or not the program will be an overall success without allowing it to run its course for a significant amount of time because unlike a plant, you can't control everything in different schools around the country. No matter what you do, there are always going to be differences in the school that could affect the success of such a program, such as how different weather might affect one's learning capabilities, how one school's financial situation might affect students' learning capabilities more than a school with another financial situation, etc.

The article also explains that through the social science experiments the author-and other-conducted, there are three lessons that have emerged:

1. Few programs can be show to work in properly randomized and replicated trials ---> one should be reluctant to believe claims of the effectiveness of new programs and policies. It is important to see how each program plays out before deciding whether or not one was a success.
2. In this universe, plans are more likely to fail than succeed. Programs that attempt to change people are more likely to fail than those than try to chance incentives or the environment in which the problem is taking place.
3. It is a rare occasion that a program creates enormous improvements---> programs that do work usually improve problems modestly in comparison to how serious the problem actually is.

The bottom line that we as human beings do not have a scientific method with which to understand human society. We may be heading that way down the line, but at this point in time, we are not yet there. The only way to learn whether or not a program is truly successful and to make a successful program is by trial and error and allowing a program significant time to run its course.