What Countries are ‘Pulling their Weight’ for Haiti?

January 26, 2010

(This article was first published on Zero Intelligence Agents » R, and kindly contributed to R-bloggers)

Using the data provided ReliefWeb on the Appeals and Funding to Haiti (h/t DataBlog) and the most recent GNP estimates, I decided to do a little “back of the envelope” analysis. With GNP as a proxy for a country’s wealth, the hypothesis is that pledges should roughly be a linear function of wealth, i.e., the more you have the more you can give. Again, this is by no means a complete, or even partially complete analysis, simply a quick exploration of the data; that said, throwing the data into the ggplot2 sausage grinder reveals some interesting features.


A few things to note from the above figure. First, the axises are log-log plots of GNP and pledge amounts in U.S. dollars. Also, country names are sized and colored by their residual from log(pledge)sim log(GNP), and absolute values are taken on size to avoid negative sizes. A red coloring indicates a positive outlier, and blue a negative. The black line is the best fit line corresponding to the above linear model.

Immediately, Norway and Sweden standout as countries that (by this very rough approximation) appear to be contributing above their wealth level, and conversely China and Japan (text overlapping) seem to be under-pledging. Also, though difficult to see, Finland, Switzerland and Spain all fall almost perfectly on the best fit line, while the United States is an extreme outlier.

I will resist editorializing this data, as there are many factors that this is not picking up; most notably, the difference between pledges and committed funds. Also, the data does not include pledges from private or international organizations, which would most certainly alter the plot. A better analysis may be to fit a gravity-like model to this data, where pledges are a function of the distance of Port-au-Prince to another country and the level of trade between Haiti and that country. Gravity models are very popular, and seem relevant to this data.

To leave a comment for the author, please follow the link and comment on his blog: Zero Intelligence Agents » R.

R-bloggers.com offers daily e-mail updates about R news and tutorials on topics such as: visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...

If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Tags: , , ,

Comments are closed.