Modern macroeconomic theory jumps between the perspective of a detached observer trying to explain what the Fed is doing, and a Fed advisor role, helping the Fed better achieve its goals.
From the first perspective, Fed behavior appears to be best explained by "interest rate smoothing." Rather than immediately adjusting interest rates to the level that is expected to most rapidly close the expected output gap and keep expected inflation on target, the Fed gradually adjusts interest rates to that level. The target Federal Funds rate appears to be a weighted average of its own past level and the level that should be expected to maintain full employment and stable inflation.
When macroeconomic theory shifts to the Fed advisor role, the Fed's goal of avoiding volatility of short term interest rates is taken as given. The Fed is trying to target inflation--the core CPI should increase at a 2 percent rate from wherever it happens to be. The Fed is concerned about the output gap, the difference between real output and productive capacity. And the Fed is concerned about volatility of short term interest rates, and "macro" is about achieving the Fed's goals.
Treating short term interest rates like inflation is a mistake. A better analogy would be the output gap. The goal should be for the market interest rate to remain equal to the natural interest rate, and so, it should be an "interest rate gap" that should be minimized. If the natural interest rate fluctuates, then so should market interest rates. Keeping market interest rates stable in the face of a fluctuating natural interest rate is a mistake. It would be a bit like trying to stabilize real output in the face of fluctuations in the productive capacity of the economy.
Interest rates are a market price. The role of market prices is to coordinate the plans of buyers and sellers. The role of interest rates is to coordinate the plans of borrowers and lenders. If it takes large changes in a market price to coordinate plans, then preventing those price changes is a disruptive interference in the market process.
Unfortunately, monetary disequilibrium--an imbalance between the quantity of money and the demand to hold money--can interfere with the ability of interest rates to coordinate the plans of borrowers and lenders. If an increase in the market interest rate occurs because the quantity of money is less than the demand to hold money, then "smoothing" or really reversing or preventing that increase in the market interest rate by an expansion in the quantity of money is appropriate. If, on the other hand, an increase in the market interest rate occurs because investors want to borrow to fund new investment opportunities, then "smoothing" that change in interest rates by expanding the quantity of money and creating monetary disequilibrium would be disruptive.
Of course, if the Fed is targeting interest rates, then there are no changes in market interest rates for the Fed to choose to reverse or not. However, a rough rule of thumb would be that a change in the output gap or inflation that the Fed might want to reverse or prevent would imply that the natural interest rate has changed, so that any interest rate smoothing would be an intentional creation of monetary disequilibrium. The Fed would be creating an imbalance between the quantity of money and the demand to hold it in order to keep a market price from changing enough to clear markets.
Incredibly, the Fed's penchant for creating monetary disequilibrium to keep short term interest rates stable is rationalized by many modern macroeconomists. The argument is plausible. (Perhaps it is correct.) A small change in interest rates, if it is expected to persist, will have a larger impact on aggregate expenditures than a larger, temporary change in interest rates. Forward looking investors and consumers will anticipate the Fed's interest rate smoothing. While the current change in interest rates may be "too small" to motivate the proper, market clearing response, market participants will understand that the Fed won't reverse this change rapidly, and so the expected persistence of the change will allow for the optimal response anyway. The result is less interest rate volatility and appropriate changes in real expenditure to reverse or prevent an output gap or changes in inflation.
Suppose that price controls are implemented in the gasoline market. Clearly, volatility in gasoline prices is undesirable, but long lines at the gas station are bad too. So, the optimal response to lines at the gas stations will be slow and measured increases in gasoline prices. Once the lines disappear, the prices should remain too high, generating surpluses for a time. Forward looking gasoline consumers will cut back on their purchases now, knowing that there will be gas in the future with no wait. Gasoline producers will sell more now, knowing that they won't be able to sell in the future when the fixed price is above equilibrium.
Perhaps a scheme of gasoline price controls with smoothing will reduce the volatility of gasoline prices and still clear markets. I doubt it.