Adam Posen and John Taylor have exchanged posts on the Fed Oversight Reform and Modernization Act. The AP criticism of the Act, in a nutshell, is that discretion is better than rules: that the Act would constrain the Fed to follow policy procedures declared in advance, that might later prove inappropriate.
JT rebuts this by pointing out that the Act does not make any rule that the Fed declares binding on itself. It simply forces the Fed to declare such a rule, and explain how it uses it and why it deviates from it.
However, JT’s rebuttal seems like pure legalism.
We know from his writings that he thinks Fed policy would have been better if it adhered to the prescriptions of a Taylor Rule, or something close to it. JT views quantitative easing as a dangerous act of discretion. It’s not clear what he makes of the international evidence that QE appeared to lower long rates on impact. But at any rate he considers the uncertainty generated by embarking on QE to have had a negative effect on activity, once that must presumably have swamped any stimulus via long rates, (and be greater than any uncertainty generated by the private sector wondering what on earth was going to counter the enormous recessionary shock, if not monetary policy.)
We also know that he supports the Act because he thinks it will make policy less discretionary than it was in the past, and more Taylor-Rule- like. A re-run of the recent financial crisis under the Act would tilt Fed policy towards what he would have preferred.
This is precisely what Posen [and Krugman, and myself for that matter] object to. Whether literally binding or not, Taylor knows that the Act will increase the chance that future FOMCs feel constrained by their prior rule declarations.
If JT didn’t consider this the effect of the Act, what possible benefit could it have? And if this is the effect the Act would have, then, if you think QE and credit easing policies, and the subsequent zero interest rate policies were stimulative and beneficial, then the criticism stands.
Taylor goes against the modern consensus that central banks should have goals given to them by Parliaments [ie have no goal independence], and be held to account for achieving those goals, but have instrument independence, and discretion to operate monetary policy as best as it sees fit to achieve those goals.
He leans on the theoretical literature which accords benefits to committing to rules for instruments. But these benefits accrue mostly because of the assumption of rational expectations, and their confinement to a fairly narrow class of DSGE models. `Even without discarding the unrealistic assumption of RE, the crisis has taught us, surely, that know much less about the other bits of the model than we thought when the commitment studies ruled.