Your Eyes Can Decieve You - Don't Trust Them

Andrew Muhlestein
19 November 2012
Political Science 150
Professor Hawkins

Blog 9: Good Governance

There has been a lot of talk about the key variables in determining whether a country will have good governance. You've asked me to look at the data of corruption and wealth and see if these are at last correlated in their relationship. In order to do so, I've chosen to use Gross National Income per capita at Purchasing Power Parity from the World Bank's data as a representation of wealth. This is a good measure because it breaks down wealth to an individual level, which leaves out much of the skewing effect of unbalanced distribution of wealth, and it levels out inflation via the purchasing power parity (so what your money will buy you, as opposed to how much you make in an arbitrary number). For corruption I've decided to use the Corruption Perceptions Index produced by Transparency International. This index is a good measure because it takes survey data from a large population sample gathering data on the number of bribes, the necessity of bribes in order to get things done, and the general feel of the people about corruption. This is the best measure as there is no reliable way to directly measure corruption, and these people actually live in the area and experience it. All data was from 2010.

After gathering the data I made a scatterplot out of it using Microsoft Excel, and these were the results.

The results showed a little correlation between GNI and Corruption. Yet this relationship is weak enough that it's difficult to say the nature of it; that is to say, whether it is linear, curved, hyperbolic, etc. It is clear, however, that if your GNI is very low, you definitely have a better chacne of having high corruption, as seen in that large clump in the bottom left corner. Having looked at my fellow... research assistants' work, however, it's clear that in other years the correlation is much stronger.

As a result of both my own findings and the findings of my fellow students, there are some conclusions we can reach about outliers. First, most countries stayed roughly the same on the list. However, a small number shift enough to form a pretty linear relationship, though their stay in that position is very temporary. It's easy to suggest a relationship when a line is formed, but we must recognize that the vast majority of the points are not part of that line, but rather form clumps. There is the low GNI high corruption clump, a looser high GNI high corruption clump, and another loose high GNI low corruption clump.

**From this information we can start to draw some tentative conclusions. First this clumping, as opposed to a linear line, suggests a different type of relationship. Instead of a linear relationship, where for example the higher your GNI the lower your corruption, we should instead look at a categorical correlation. That is to say, if your country fits in these categories, then you will probably be in this clump. This lends a lot of credence to the colonialization theories, where ex-colonies form one of the clumps.  This would be because of the host of similarities between those countries, such as type of installed government, shared (roughly) history, similiar language, stronger trade relationships between them, similar justice systems, similiar tariffs/trade laws, etc. There's a lot of room within that for different clumps, such as whose colony you were, etc., but the approach is nonetheless solid.

Sources

"Databank User Session Time Out." World Databank. N.p., n.d. Web. 19 Nov. 2012. <http://databank.worldbank.org/ddp/home.do?Step=3&id=4>.

"2010 Corruption Perceptions Index -- Results." Transparency International - the global coalition against corruption. N.p., n.d. Web. 19 Nov. 2012. <http://www.transparency.org/cpi2010/results>.