The article that I read was one from the New York Daily News from September of this past year. It discusses how poverty rates in New York City rose according to the 2010 census. They state that the economy is to blame for the rise in poverty rates and that there is concern of how poverty is spread throughout the city. There is a lot of disparity in poverty among different demographics of people. The article also states that every borough except for Manhattan experienced an increase in poverty rates which is also concerning to government officials in the City.
The article made me realize that something that I may need to keep in mind are years in which there are recessions and how it may affect the poverty rate. Obviously, if the economy is not doing well then poverty rates are more likely to go up in those years. Therefore, I could possibly create a dummy variable for years that are and are not recession years. This way I could hopefully control for the economy and get a better measure of how population is effecting poverty rates.
Another thing the article talks about is poverty among different racial groups. This is something that I could maybe think about as well. I am not exactly sure how I would take that into account or whether it would skew my data at all if I don't consider it, but it would definitely be something interesting to explore. I don't think that any of these new variables would violate any of the assumptions of the linear model however, it is possible that there may be correlation among variables that I choose to use. There is likely correlation between recession years and the poverty rates which implies co-linearity among my explanatory variables. However, I think including information about the economy, specifically the recession years as dummy variables will help to give me a stronger regression.
No comments:
Post a Comment