Crystal balls and P-values
I wanted to take time out to talk about something that is getting me pretty hot under the collar these days and it’s not what you would expect. Well actually maybe you would expect it, because by and large I come across as an even bigger nerd in this blog than I even am, which I am starting to realise is quite a feat, but still…here goes:
Remember how I said that Econometrics was my most challenging class this quarter? Well that’s still the case. Econometrics is defined as “the application of mathematics and statistical methods to economic data” and is described as the branch of economics “that aims to give empirical content to economic relations.” You begin to get the picture, right? Art class this aint (and that’s coming from someone who took Art up until my final year in school….Including a 12 hour final “painting” exam! Ha ha ha…)
Thing is, now that we’re further along in the class I am beginning to see the applicability and utility of this discipline and the software that goes with it (STATA) such that I have to say I am actually pretty genuinely excited about using this more in my studies and potentially in a future career.
That’s not to say I don’t also respect the arguments from people like Professor Deirdre McCloskey (see earlier blog post regarding the transgender, libertarian economist who came to speak at Korbel earlier this year) who say that this discipline is dangerous if the output of these mathematical contortions is applied without due consideration of the real life, economic and social impacts in society. But as someone who feels that policy making WITHOUT adequate analysis of related data (e.g. trying to solve health, education or crime-related problems without looking at what factor into those trends from a quantitative perspective) is likely to be futile if not outright negligent… the realisation of the greater level of understanding and insight we can obtain about an issue through econometric analysis is really encouraging.
For example, The Bahamas just had an election last Monday. The incumbent government lost badly, winning only 9 seats to the new Government’s 29. Among other things, spiraling crime rates are being blamed. The murder record each year for the last five years has broken the previous year’s record. The question is, what can be done to tackle this issue?
I’m hoping I might be able to get an insight into this issue in my econometrics class of all places. I plan to run an econometric analysis where I will determine the relationship between factors such as tourism, government spending on police and perhaps other areas, unemployment rates, foreign direct investment inflows and the violent crime rate across a series of years. In doing so, I expect to see which of these factors has the greatest impact on decreasing the crime rate – – for example, is it simply whether you have a job that matters, or if the government is spending more on the police to control bad behaviour? If so, what impact would increasing government spending on police have on the violent crime rate? (Here’s where the crystal ball comes into it…).
Anyway, I really hope I come up with something significant in my findings and can make some useful projections about what kind of policies might flow from that to try to bring down the crime rate. Or at the very least, some reasonable recommendations on further research from which I or someone else could get an even better insight into how we can begin to tackle this scourge. Or I might choose the wrong varibales, totally misinterpret my output and fail miserably…fingers crossed it’s not the latter. Either way, it’s a learning process.
On an unrelated note, I just discovered that wordpress now compiles information on which countries those who have viewed this blog were located in. Turns out since February 25, 2012, people in 54 different countries have seen the blog. I think in some cases it will come up when people are just searching certain keywords or images, but still pretty cool to think of my rambles popping up on screens across the globe. Greetings!