Michael S. Hanson Department of Economics

HOME

TEACHING

RESEARCH

LINKS

OTHER

Economics 300: Problem Set #8

The following problem requires the use of EViews, a statistical package available on the PCs in the PAC Lab and most other labs on campus. If you need help with EViews, you should ask the TAs for assistance, consult the built-in help from within EViews, and/or browse one of the "tutorials" listed on the Econ 300 Resources Page. I also may be able to answer questions during office hours. Inflation and Money Growth In EViews, create a new workfile with an annual frequency, starting in 1959 and ending in 2000. Then load the following series from the USECON database into EViews: PDGDP: Gross Domestic Product: Implicit Price Deflator (SA, %Chg) FM1: Money Stock: M1 (SA, Bil.$) Hint: The USEcon database is located in the following directory: S:\DATA_ANALYSIS\EVIEWS\EVIEWS4NET\DATA\USECON.EDB. The easiest way to access this database may be by clicking on the [FETCH] button in the workfile window. The TAs may illustrate other ways to access this database. Hint: Don't forget to save your workfile! We may use this workfile in another problem set in a week or two. Generate a new series for the growth rate of the M1 money supply and save it in your workfile. Hint: The units should be percentage points, just as they are for the growth rate ("% Chg") of the implicit price deflator. That is, 3 percent annual growth should be expressed as 3, not 0.03. Hint: As we will show later in the course, the growth rate of a variable X can be well approximated by the one-period difference in the logarithm of X. EViews has a built-in function for this common calculation: dlog(X). Hint: To confirm that you have calculated the growth rate correctly, you should check your calculations for the past few years (say 1998 - 2000) against an official source, such as the Economic Report of the President or any number of publications at the Federal Reserve Board. Plot a line graph with both the inflation rate (the growth rate of the GDP implicit price deflator) and the growth rate of money. Print this graph. Comment upon the apparent relationship between these two variables over time, and how it appears to change, if at all. Hint: For this question and several of those that follow, you may find it useful to create a "group" object in EViews. First, click once on your money growth variable to select it. Then control-click (once) on PDGDP to select it as well. Next, right-click on either series name and select Open > as Group  from the pop-up menu. In the resulting window, click on the [View] button to see the various options (including graphing) available to you. Hint: Double-click on the graph to adjust the plotting options. You may find it useful to add a zero line for the x-axis and an informative title, for example. The purpose of a graph is to clarify exposition, so avoid the temptation to add fancy stuff that defeats this purpose. Compute the correlation between inflation and the money growth rate for the full sample period, 1960 – 2000. Compare with the correlation over the 1960 to 1981 period, and with the correlation over the 1982 and 2000 period. Use figure 15-4 in the textbook to construct approximate confidence intervals for the true correlation in each sub-sample. Report these intervals; do they overlap? Hint: Click on the [Sample] button to adjust the sample period in effect for your particular statistic. Don't forget to change it back when appropriate! Graph a scatterplot with the growth rate of money on the X axis and the inflation rate on the Y axis. Print this graph. Does the relation shown seem consistent with your results in the previous two questions? Explain. Now graph and print separate scatterplots for the 1960 – 1981 and 1982 – 2000 sub-samples. Does the relationship between these two variables appear to be the same in each of these two periods? Explain. Hint: Double click on the graph to add descriptive labels and titles, so you can distinguish one sub-sample from another. For the full 1960 – 2000 sample period, regress inflation on the growth rate of money (and a constant, of course!) Print out the results. Report and interpret the estimated value of the slope coefficient. Is it statistically dicernible from zero? From one? Explain, showing your computations. Hint: There are numerous ways to tell EViews to estimate a regression of Y on X: Click on Quick in the main menu of EViews, then select Estimate Equation.... Up will pop a window (called Equation Specification) in which you can tell EViews what you want to estimate. In the large box, type "Y C X" to estimate a Least Squares regression of Y on X. Click [ OK ]. From your workfile, select your Y variable and then your X variable with the mouse so that both are highlighted. Right click on one of these series and select Open > As Equation... from the pop-up menu. If the variables are in the correct order, i.e. listed as Y X C, then click [ OK ]. If not, edit the list of variables in the window and then click [ OK ]. If you already have the two series selected as a group (because you calculated their correlation or graphed a scatterplot, say), you can click on the [Procs] button at the top of the group window, then select Make Equation... from the menu. Make sure the variables are listed as "Y X C", then click [ OK ]. The long white bar just below the main menu of EViews is the "command line." Place the cursor in that space and type "LS Y C X" to estimate a Least Squares regression of Y on X. Hint: Always, always, always include a constant (i.e. intercept) in your regressions. EViews denotes the constant as C. Be sure to include it in your regressor list! (Notice that it doesn't matter if the C comes before or after the X -- all that matters is that both follow the Y in the above expressions.) Hint: Don't forget to change the sample back before answering this question! Repeat the same regression and answer the same questions as above, for each of the two sub-samples we have been considering. What happened in the early 1980s that explains these results? Hint: What does macroeconomic theory suggest should be the long-run relationship between prices and money? Compare the correlations between inflation and money growth with the R-squared values for each of the three time periods studied (the full sample plus the two sub-sample periods) Do you see a specific relationship? Explain. Comments: You may discuss this problem set with your classmates, but do your own work. Submitting any part of another student's work as your own is a violation of the Honor Code and is not acceptable. Print your name and Wes ID number on each page you submit. You should place all of the written answers on one page. (If you want to place more than one graph on a page, you may do that as well.) While this problem is not that hard once you get going, if EViews (or, heaven forbid, introductory macroeconomics!) is unfamiliar to you, starting Monday evening is a recipe for no sleep before Tuesday's class. Please plan accordingly!