Monetary Policy: Work it Out

In the United States, the governments data on real growth improve over time. For instance, we now know that in the early 1970s, the economy was actually growing 4% faster than people believed. Economists at the Philadelphia Federal Reserve have collected data on how our view of the economy has changed over time. Lets use their summary of the data and a virtual sixsided die to see just how inaccurate our realtime views of the economy actually are.

a. We are going to reenact the 1970s, and we will start figuring out how errorfilled the governments growth estimates will be. Croushore and Stark report that, on average:

i. Onesixth of the time, measured growth is 2% better than actual growth.i. Onesixth of the time, measured growth is 2% better than actual growth.

ii. Onethird of the time, measured growth is 1% better than actual growth.ii. Onethird of the time, measured growth is 1% better than actual growth.

iii. Onethird of the time, measured growth is 1.5% worse than actual growth.iii. Onethird of the time, measured growth is 1.5% worse than actual growth.

iv. Onesixth of the time, measured growth is 3% worse than actual growth.iv. Onesixth of the time, measured growth is 3% worse than actual growth.

Pretend you have rolled a sixsided die to simulate the government growth estimates for a fouryear period in the 1970s. The results are as follows:

1974:31974:3

1975:21975:2

1976:21976:2

1977:51977:5

Record your rolls in a table. If you have rolled a one, measured growth in that year is 2% better than actual growth. If you rolled a two or three, measured growth in that year is 1% better than actual growth. If you rolled a four or five, it means measured growth in that year is 1.5% worse than actual growth. If you roll a 6, measured growth in that year is 3% worse than actual growth. Write down how much measurement error you will have for each year.

Example: If your first roll was a 4, that places you in category iii, so write down -1.5 as the amount of measurement error for 1971.

b. Lets see what values we get when we add together the true real growth rate with the measurement error in the previous table. For true real growth, we use the most recent data in the following tablebut of course even these estimates could change in the future. The sum is the actual government data that will wind up in the Federal Reserve chairs hands.

Example: Consider the year 1971. If your first roll was a 4, that placed you in category iii. Therefore, you subtract 1.5% from the true 1971 growth rate to yield a realtime government report of 1.9% annual growth. Here are the actual growth rates:

1974:0.5%1974:0.5%

1975:0.2%1975:0.2%

1976:5.3%1976:5.3%

1977:4.6%1977:4.6%

c. In your simulation, how many times was the government data off by 2% or more?