252CP4-08 4/18/08

Computer Problem 4

This is an internet project as mandated by the Economics/Finance Department. Possible credit is 2 to 8 points depending on effort and originality. It is due as part of the final exam.

Do only one of the following 3 problems. The most common errors in this sort of material is not knowing what you are trying to explain. The variable that you are trying to explain is your independent (Y) variable and the rest are independent. Turning in equations with lots of insignificant coefficients isn’t very useful either.

Problem1) The State of Working America has a graph relating life expectancy at birth to per capita health expenditures in OECD countriesin Chapter 8 This material is beginning to get a bit old. (The URL of the graph is

.

The reason that I picked this particular graph is that life expectancy is still one of the most accessible measures of human welfare and health expenditure is an obvious determinant, though not necessarily the mot important. Please do not attempt to read values of Life expectancy or Health Spending from this graph.Get them from a website. The relationship suggested by this graph is decidedly nonlinear, so that you might want to try, as your first independent variables, health spending and health spending squared. (If health spending is in C2, you can put health spending squared in C3 by using the Minitab command ‘Let C3 = C2 * C2’). Or you could try the logarithm of health spending is place of these two (‘Let C3 = Loge(C2)’). Thedependent variable would be Life Expectancy and other possible independent variables could be things like per capita income, unemployment rates, poverty rates and the Indexof Economic Freedom. Do not use independent variables that would reflect the size of the country like total GDP or total spending.You are, of course, looking for significant coefficients and high R-squares. For instructions on how to do a regression, try the material in Doing a Regression and in the problem solution suggested there.See also the first question on last year’s final exam. I expect some commentary, using R-squares and significance, on what independent variables are good predictors. I would be particularly interested in the relationship of Life Expectancy to Economic Freedom. Limit yourself to the OECD countries mentioned in the chapter that this diagram comes from (unless, of course, you have good reasons not to). There are some statistics there that can be copied. Other sites that I looked at are below.

You can find updated life expectancy here.

OECD sources give updated life expectancies and heath expenditure. The most current version seems to be below.

This is the CIA Factbook, the part listed under text version seems relatively easy to read. It has been updated for 2008.

This is the source for the Index of Economic Freedom. What it tracks may be indicated by the fact that the US position fell last year because of the rise of the minimum wage. Nevertheless, it would be interesting to see if its components work well to explain various measures of health.

Wikipedia is not usually considered a reliable source, but these seem to come from reliable sources.

Their ‘Economic Freedom of the World’ data sets look a bit old to me.

This is very current and has a ‘build your own table’ feature. Its human development index might be an alternative to life expectancy as a dependent variable.

This gives relative poverty rates for the entire population, children and the elderly. It has some date problems and you should make sure you know what relative poverty means before you use this stuff.

This is the most recent World Bank data site and seems to have a ton of stuff.

. A book on the happy planet index is available on-line it or, better, its components are possible independent variables. I didn’t have tome to figure out exactly how they computed the index, but the ‘in your face’ part is that the US doesn’t do very well.

Sites that we have used in the past are below

But just by wandering around the clippings on my website I found many other potential sources. Here is my list with the total zeroes removed.

A listing of many worldwide sources of economic data.

Lots of fancy graphics here, but it may take some work to find numbers. They seem to be in

A large set of data limited to failed states. It could be interesting to relate this to economic freedom.

Data and components for many nations.

A site everyone should go to, but it may take ingenuity or a subscription to get much hard data. I’ve posted some price data from it at

Crime victimization rates

This site is not exactly user-friendly, but with a little effort you can get Excel spreadsheets associated with their maps with GDP/capita and lots of other good stuff.

Worldwide rankings on the ease of doing business.

There is a corruption perceptions index in here.

Problem 2)Try to explain some measure of firm equity performance, perhaps price-earnings ratio or percent gain in stock price over 1 to five years by doing a multiple regression against other variables that might describe a firm. These independent variables could include earnings per share, rate of change of earnings per share, net profit/net sales, current ratio, % gain in revenues (over 5 years?), CEO salary and industry dummy variables. The industry dummies are necessary because market interest is likely to work differently in different industries. You should limit yourself to no more than 30 firms in no more than four industries. Possibly useful websites follow and others may be added if other faculty members suggest them. You are, of course, looking for significant coefficients and high R-squares. Dr. Snow is particularly interested to see if there is a relationship between the dependent variable and CEO salary. For instructions on how to do a regression, try the material in Doing a Regression and in the problem solution suggested there.See also the first question on last year’s final exam. I expect some commentary, using R-squares and significance, on what independent variables are good predictors

2008 data has not been posted.

- for definitions of terms.

Lots of info on many company stocks.

Problem 3) Propose a problem using internet data and multiple regression and have it approved,