The aim of Gramm-Rudman (GR) was to achieve a target value of the federal government deficit each year. If the target for a given year were not going to be met, then Congress would be required either to raise taxes or to cut spending (or to do some of both). As a last resort, Congress would pass "automatic" spending cuts, where most of the items in the budget would be cut across the board by the amount necessary to meet the deficit target.

From a macroeconomic perspective, what GR did was to make the deficit an exogenous variable. The deficit variable in the model is SGP. In the regular version of the model SGP is endogenous. It depends on the tax revenue of the government, which is endogenous because tax bases are endogenous (tax bases depend on the state of the economy). SGP also depends on some government expenditure items that are endogenous. If SGP is taken to be exogenous, then some variable that was exogenous before must be taken to be endogenous. (With the same number of equations, we must have the same number of unknowns.) The software allows you to choose one of three variables to take as endogenous when SGP is taken to be exogenous: government purchases of goods (COG), the personal income tax parameter (D1G), and transfer payments (TRGH).

The first three experiments in this chapter are concerned with cutting the deficit by $20 billion. For the first experiment COG is taken to be endogenous, i.e., to be the variable that adjusts to achieve the deficit target. For the second experiment D1G is taken to be endogenous, and for the third experiment TRGH is taken to be endogenous.

Note that -SGP, the government deficit, is in current dollars, while COG, government purchases of goods, is in constant 1992 dollars. In order to examine what happens to government purchases of goods in current dollars, you must multiply COG by PG, where PG is the price price index for COG. Thus, to see how much government purchases in current dollars must be cut in order to achieve a deficit reduction, i.e. a reduction in -SGP, you must multiply COG by PG. Then, both the deficit and government purchases will be in current dollars, and you will be able to make valid comparisons.

Experiment 8.1: Decrease in the Deficit; COG Adjusts

• Cut -SGP by $20 billion for the forecast period, and take COG to be endogenous. Solve the model.

1. By how much did COG*PG, government purchases in current dollars, have to decrease to achieve the deficit reduction of $20 billion? Why is this amount more than $20 billion? (Note: the answer to this question seemed completely lost on Congress, which seemed to think that to cut the deficit by, say, $20 billion, you only need to cut spending by $20 billion.)
2. What happened to real GDP, the unemployment rate, and other measures of economic activity?
3. What did the Fed do in response to the deficit reduction? Do you think that in practice the Fed would respond more than the model predicts it would? How much more? (Note: you can experiment with alternative monetary policy assumptions if you wish. In other words, you can drop equation 30 and instead choose values for RS or M1.)
4. What happened to the interest payments of the federal government INTG? Did this help or hurt in achieving the deficit reduction?

Experiment 8.2: Decrease in the Deficit; D1G Adjusts

• Cut -SGP by $20 billion for the forecast period, and take D1G to be endogenous. Solve the model.

1. Why is the fall in real GDP less in this experiment than it is in Experiment 8.1?
2. Compare the differences in the labor force responses between this experiment and Experiment 8.1.

• Cut -SGP by $20 billion for the forecast period, and take TRGH to be endogenous. Solve the model.

1. Why did TRGH have to fall by more than $20 billion? (TRGH is current dollars, so you do not have any conversion problems here.)
2. Compare the results of this experiment with those of Experiments 8.1 and 8.2. Focus on real GDP differences and differences in labor force responses.

We have so far seen that government spending must be cut by more than the targeted drop in the deficit. (You should have a good understanding of why this is so.) We next consider how the economy differs in its response to shocks under GR. We will see that GR is in fact destabilizing. We consider a $20 billion decrease in exports (caused, say, by events abroad that the United States has no control over).

Experiment 8.4: Decrease in Exports; no Gramm-Rudman

• Decrease EX by $20 billion in each quarter of the forecast period. Solve the model.

1. How much did the deficit increase as a result of this negative export shock? Why? What happened to real GDP?
2. What did the Fed do in response to the shock? In particular, what happened to RS?

Experiment 8.5: Decrease in Exports; Gramm-Rudman in Effect; COG Adjusts

• Decrease EX by $20 billion in each quarter of the forecast period. Take SGP to be exogenous, where COG adjusts. Solve the model.

1. What did COG do in response to the export decrease and why? Remember, for current dollar COG, use PG*COG.
2. Why is the fall in real GDP larger in this experiment than it is in Experiment 8.4? Compare carefully the differences between this experiment and Experiment 8.4. You should have a good understanding of why Gramm-Rudman is destabilizing.
3. What did the Fed do differently here than in Experiment 8.4? Compare the values of RS and AG in the two experiments.

We now consider a price shock and examine how the economy behaves differently under Gramm-Rudman than otherwise. In Experiment 7.1 the import price index (PIM) was increased by 10 percent. We now consider the same experiment except that we assume that GR is in effect with COG adjusting.

Experiment 8.6: Increase in PIM; Gramm-Rudman in Effect; COG Adjusts

• Increase PIM by 10 percent in each quarter of the forecast period. Take SGP to be exogenous, where COG adjusts. Solve the model.

1. What did COG do in response to the price shock and why?
2. Why was the change in GDPR different here than it was in Experiment 7.1?
3. Is Gramm-Rudman stabilizing or destabilizing in the case of price shocks? Why?

Now that you are familiar with running Gramm-Rudman experiments, you may want to run others. You should note that it is possible in the program to use combinations of tax increases and spending cuts to achieve a given deficit reduction. Say that you want to decrease -SGP by $20 billion by partly raising D1G, partly decreasing TRGH, and partly decreasing COG. You choose ahead of time how much you want to raise D1G and lower TRGH. You then enter these changes along with telling the program that you want SGP to be lowered by $20 billion with COG adjusting. When you then solve the model with these changes, COG will only adjust by the amount that is needed to reach the deficit reduction after taking into account the increase in D1G and the decrease in TRGH.

You can also have the deficit reductions change over time. You may, for example, want -SGP to decrease (from the base values) by $20 billion in the first year, $40 billion in the second year, $60 billion in the third year, and so on until the deficit is eliminated.

Note also that SGP is the surplus or deficit of the federal government on the NIPA basis. Most of the time the Gramm-Rudman deficit targets quoted in the press are for the deficit of the unified budget. The NIPA and unified budgets differ somewhat, and you will have to figure out how to translate unified budget targets into NIPA budget targets if you want to be precise.