Revised August, 2004

This is a simple exercise for the Fairmodel, which you can do yourself on your home computer, for free. Or, I am glad to do it for you in my office. In this exercise, we examine the effect on real GDP of an increase in government expenditures of $10 billion. We will see that the multiplier reaches a value of 1.5, and then declines back to about unity.

From my Webpage, or the Economics Page, go to the “Invitation to the Fair Model,” and from there click on--Fairmodel (http://fairmodel.econ.yale.edu)

On the left hand side of the next screen, click on "US Model". On the next sheet, scroll down, and click on “Solve.”

On the screen that comes up, you have to make up a name and password which will identify the dataset (these will have some importance with later exercises, but not now. They erase the old files every couple of weeks). Give your file a name, and a password, and click [Enter]. Click on "Copy Base" which allows you to compare your new scenario to the baseline scenario. The machine then tells you on the next screen if everything was accepted; if it was, just click on [Proceed].

Now we should be on the major screen, which is titled "Start.html". From this page we give the major instructions to this software. The first two steps simply confirm default configurations.

With regard to the time period for the prediction, we will use the default of 20043 to 20084 (meaning third quarter of 2004 thru fourth quarter of 2008); so click on [Save Changes].

On the next screen that comes up, click on "change assumptions on monetary policy." (This happens to be a major contribution of the Fairmodel, and is discussed in detail in a later exercise). For the purposes of this intro, stay with the Regular version of the model, and just hit [Save Changes].

Once again you're back at the "Start.html" page, and you want to click on "Change Exogenous Variables."

The machine comes back to you with a divided screen. On the right hand side, click on COG (we're going to increase government purchases by $10 billion). The right side of the screen now shows the data for COG; if you'd like, you might scroll down to see what the data look like. Now, go to the left side. Click into the "Add to each of the existing values", then type in 10, and hit your [return] button. You might look over at the right hand partition on the screen, and scroll down to check that your data has in fact increased by 10 for the prediction period. If this is the case, then go back to the left hand side, scroll down, click on [Commit to Change]. Then, scroll back up to the top of that left hand screen, and click on [Home]. You should now be back again to the "Start.html" page. Because this is a real simple exercise, that's the only change in the variables that will be imposed. So, scroll down the "Start.html" page, and hit "Solve the model and examine the results" (which currently is option #8).

The machine now solves the model, and tells you how many iterations were requried for each quarter. If it takes each round only two or three iterations, then you're probably wrong. If there are 10-20, then it's ok. Scroll down that page, and hit [Proceed]

Now we're looking at the "outframe.html" screen, and we're almost ready to have the machine tell us what our results are. You can choose that the presentation be in graphs or tables, variable levels, etc. I always prefer to compare to the Base, and so on "Comparison Dataset" click on "Use Base." In the bottom of the page, in the "List of Variables for Output" click on "GDPR, GDPD, RS, M1, UR, SGP" (which are real gdp, the gdp deflator, short term interest rate, and the rate of unemployment, and the government's surplus/deficit). Having highlighted these, click on "Add to Current List." Leave everything else at deafualt, and hit [Display Output]. The next screen is your output, showing the "Base" and the new levels. Here is a copy of the actual output screen.

On the output screen, you can see that real GDP goes up by about 16, and then slides back -- compared to BASE. Recall that in this exercise we increased government expenditures by 10. The interpretation is that there is a short period during which the multiplier is about 1.6 (16/10), then it quickly falls back to about one. Not much change in unemployment or interest rates, but $10 billion isn't such a big kick in an economy of a $2,000 billion GDP.

It's not so hard. The Workbook is rather clearly written, and the exercises are useful.

Click here to return to the independent study course announcement, or the syllabus.