Suppose you have two objectives for monetary policy: financial sector stability and real economy stability. You have two main instruments: interest rates and macroprudential policy (things like changing the capital requirements of banks, or making it less easy to get a mortgage; often called macropru for short). Should you assign one instrument to one objective (often called an assignment), or try and achieve both goals with both instruments?
As Tony Yates points out, assignments are rarely optimal from a purely macro point of view. Even if, say, interest rates are less effective at achieving financial stability than macropru, you would still want interest rates to contribute something to that objective. An exception is when one instrument completely dominates the other: then assignment is optimal. An example of this exception in a certain class of model is monetary over fiscal policy as means to achieve real stabilisation, as discussed here and here (less technical discussion here). But this type of result is unusual, and even when you get a result like this for a certain class of models it is not hard to add real world complications that remove the result.
If assignment in the case of real and financial stabilisation is not optimal, does this imply that those setting interest rates should take financial stability into account? It is important to understand what we are talking about here. We are not talking about how financial conditions might influence what interest rates need to be to achieve real stabilisation, which is the kind of issue discussed in this post by Bianca De Paoli for example. Nor is it talking about allowing for risk as I discussed here. Instead it would be saying that institutions like the US Fed should have a triple mandate when setting interest rates, where the third mandate was financial stability. Interest rates could be changed if financial stability was a concern, even if this had no implications for inflation or output. Equivalently, interest rates could move to help financial stability even if this took output and inflation away from target.
However we already have assignments that are clearly suboptimal from a purely macro perspective. Fiscal policy is assigned to the control of government debt, and reducing debt is certainly not a monetary policy objective. But in standard models this assignment is clearly suboptimal: when debt is high, reducing interest rates can be quite effective in reducing debt (particularly if government debt is mostly short term), and undesirable knock on effects on output and inflation can be countered by fiscal policy. Yet the mainstream consensus is that monetary policy should not be used to reduce debt.
The reason for this suboptimal assignment is most probably because of political economy concerns. Or to put it another way policy is mainly concerned about knowingly sub-optimal decisions. Reducing interest rates to reduce debt sounds too much like fiscal dominance. There is a fear that if there is no institutional assignment, politicians will depart from optimal policy and by keeping rates too low to reduce debt they will allow excessive inflation.
Can such political economy concerns be applied to financial stability and monetary policy, particularly as the same actor - an independent central bank - is in charge of both? My instinct is that it can, because central bankers are heavily influenced by pressures from the financial sector. As a result, there is a danger that if interest rates are set with both objectives in mind, central bankers will depart from optimal policy and, for example, needlessly raise interest rates before the real economy requires them to rise. The recent experience of Sweden is a clear example where this happened. For this reason I rather like the UK institutional set-up, with a separate MPC and FPC and a clear assignment for each.