At the end of this term, everyone in the Master's program has to write a paper for our econometrics class. The paper can be about anything, which is one of the reasons I'm having a lot of trouble coming up with a topic. The one requirement is that the paper has to use multiple regression - so, basically, you need to create an equation where one variable is affected by many others. The baseball player model I gave last week is a good example - the value of a player can be measured by the income he generates, so you need to find out how many games that player wins for his team, and then find the revenue from how many more tickets are sold as a result of those wins. It makes sense, and the numbers actually work out. Those are two more limiting features for this paper - although these are not necessary features, I don't want to spend 60 hours on a paper that proves nothing. People often say, "A proof of no relation can be interesting in itself", but I'm not one of those people.
So with that in mind, I have partially decided to do a paper that involves numbers that are almost guaranteed not to work out. And when I discussed this idea with a professor, he basically told me that I probably will not find any relationship. But I'm kind of set on it because I haven't been able to come up with anything else. The focus of this paper will be on remittances, which are funds that are sent from one country to another, person to person. They are not used to buy a good or service - just a transfer of money, like when your grandmother sends you a birthday check in the mail. Remittances may be mailed, as in this example, or wired through Western Union, or just physically brought by a traveller. Many remittances are sent by people who have left a poor country, like Guatemala, to work in a rich one, like the US. Over the past few years, the value of these remittances has increased dramatically, due to the ease of sending money electronically, the housing boom in the US, and other factors. In some countries, remittances account for around 10% of GDP, including Jamaica, Nicaragua, the Dominican Republic, and El Salvador. This is a pretty significant amount of cash flooding into a country, and I think that it will have an impact on their economy. The question is, what is this impact? And do I have the data necessary to show this?
My first thought is that inflation will have to go up. With all this money coming in, people will buy more, so prices will have to go up. Unfortunately, inflation is a very spastic number - especially in the developing countries that tend to receive a lot of remittances. In 1992, inflation in Jamaica was around 77%, then dropped to 10% five years later. As a comparison, inflation in the US in the past 15 years has varied from 5% to 1%. In addition, there is no simple formula for inflation - it almost always depends on consumer expectations, which are impossible to measure. This may not be a horrible idea - but the data looks pretty ugly. My next thought was to say that remittances will decrease poverty - you could say that they're a form of international welfare, considering that most people don't send money to rich families. But here poverty is difficult to measure, especially if we're using tax returns to find the poor, the most common method. For one, truly poor people don't file tax returns. For another, these returns won't include any remittances received. It would be nice to see a relationship here, but it's unlikely.
So those are my ideas, thus far. I'm probably going to try both, see if either works, and if not - bang my head against the wall. If anyone out there has any comments, or suggestions for this paper, let me know. kaifilion at gmail.com (I'm scared of the spambots, I already get 30 msgs or so a day).