36-402, Data Analysis, Spring 2011 Due 25 January 2011
Many theories of economic growth say that it’s easier for poor countries to grow faster than rich countries — “catching up”, or the “advantages of backwardness”. One argument for this is that poor countries can grow by copying existing, successful technologies and ways of doing business from rich ones. But rich countries are already using those technologies, so they can only grow by finding new ones, and copying is faster than innovation. So, all else being equal, poor countries should grow faster than rich ones. One way to check this is to look at how growth rates are related to other economic variables. We will use the np package on CRAN to do kernel regression.1 Install it, and load its oecdpanel data set. This contains growth data for many countries for 1960–1995, collected by the Organization for Economic Cooperation and Development (the OECD). We won’t use all the variables this time. GDP is “gross domestic product”, the total value of all economic production. It’s usually reported per capita and per year. Call it Yi,t , since it depends on the country i and the year t. GDP isn’t perfect2 , but it is standard. In oecdpanel, the variable growth is the logarithmic growth rate of GDP, = log Yi,t+1 /Yi,t . We look at logarithms because economic models suggest that the factors which affect growth should multiply together, rather than adding. What’s actually recorded here is the average growth rate over a five-year period, reducing year-to-year accidents. initgdp is log Yi,t , the logarithm of per-capita GDP at the start of each five-year period. A country’s investment rate is the fraction of its GDP that goes into building or repairing productive assets