In the previous post on incorporating more variables in the effective demand equation, I ended up with this graph.
There are still gaps that have to be filled. Let me point them out.
If I increased the coefficient on the variable for monetary policy (ED-FF), I could close the gaps at 1981 and 2000, but the gap at 1979 would open a lot. Like this...
Would it be possible to close those gaps in graph #2 without opening up gaps that have already closed? Would it be impossible or super complicated?
Well, let's go back and look at the stagflation of the late 70's. What happened? The abrupt rise in commodity prices produced a price supply shock which affected demand to the extent that production had to be cut. As the effective demand model is designed to show the times when production is cut, it has to incorporate the effects of supply shocks too.
Let me first place the new graph based on new changes to the model.
You will see that the huge gap in 1981 was closed while the gap in 1979 did not open. Closing that gap in 1981 is a big accomplishment in understanding the effective demand limit.
You will also see that the gap in 2000 closed. This also was an important gap to close in order to mark the beginning of the 2001 recession. You will also see that the gap from 1995 to 1998 was very stable. Maintaining the stability of this gap in the model was also important because the economy was riding the natural limit of real GDP in a stable way during those years.
We can also see that effective demand was biting into production in 1988 setting up the eventual recession which started in 1990.
So, what did I do? I incorporated a variable for supply shocks that trigger stagflation... The variable is the Consumer Price Index of all items, yoy%. (link) Simply put, I incorporated inflation. The idea was to see how much a rise in prices to consumers would affect the effective demand limit.
We see large jumps in the 70's and smaller jumps later. The jumps in the 70's could close that gap that opened up in 1979. The idea is that a quick jump in prices will take the consumer by surprise to the extent that they will lower their demand for production. If this jump happens at the right time and in the right way, it can lower effective demand to the point where it falls upon expanding production and the business cycle will stop expanding.
It seems the large overshoot of effective demand in 1973 reflects the extent to which the economy was taken by surprise by higher commodity prices, such as fertilizer from Peru and oil from OPEC. Inflation began to rise briskly due to that overshoot. Production was not cut soon enough. And even though there was another supply shock later in the 70's, we do not see an overshoot of effective demand due to a combination of labor share lifting up and monetary policy being much more accommodative.
The equation of the limit function in the new graph is this...
Limit function, L = 0.765LS + 10NX/rGDP - 55CPIall + 80(ED-FF)
This limit function (L) is then entered in the effective demand equation...
Effective demand limit, EDL = rGDP*3*T/L (1 - 0.667*T/L)
T = TFUR, capacity utilization * (1 - unemployment rate)
How can you do this equation at home? You first set the coefficients for the variables LS, NX/rGDP and CPIall. This gives you a baseline for the limit function. Then you add this baseline into the Effective Demand monetary rule (ED) to evaluate the reaction function of monetary policy (FF) to the baseline economic conditions. Then you feed back (ED-FF) into your limit function and adjust for the coefficient on (ED-FF).
So you have to be able to calculate a monetary rule rate using the limit function (L) and then compare it to the actual Fed rate. Since I am the only economist currently with a monetary rule based on a limit function to the utilization of labor and capital (TFUR), an economist would have to use my Effective Demand Monetary rule to crunch the numbers for this model. They would not work in the Taylor rule, unless it was modifed.
At this point, this new equation is saying that there is currently spare capacity still availabe to the economy and that the unemployment rate would go down to 5.5%. When we eventually see the top of this business cycle, we will have another data point to refine the coefficients in the limit function.
One final note... Look at how production overshot the effective demand limit between 2005 and 2007. It should not be that tight, otherwise a recession would have occured earlier. The reason is that long term interest rates did not rise in unison with the rise in the overnight Fed rate. The reason normally given is the foreign funds flooding back into the US from countries like China. So monetary policy was not as tight as the limit function is assuming. So the variable (ED-FF) must develop a way to incorporate the differential between short-term and long-term interest rates.
The equation is not be perfect yet, but it is getting better.