At the recent Jackson Hole meetings, John Williams, president of the Federal Reserve Bank of San Francisco, argued there has been a “significant decline in the natural rate of interest, or r* (r-star), over the past quarter-century to historically low levels.” Regular readers of this blog might recognize the view, as I have expressed it on more than one occasion. James Mackintosh follows up at the Wall Street Journal by noting that our estimates of the natural, or equilibrium, rate are too broad to draw such a conclusion. Pointedly, he states that the margin of error “is big enough to drive a truckload of economists through.” For what it is worth, I think we can be reasonably confident that the natural rate has indeed fallen. Robert Gordon’s work makes a strong case that total factor productivity has declined, which would push the natural rate down. Nonetheless, Mackintosh is correct to note that our estimates of the natural rate are very poor. Even if we were to agree that the natural rate is lower today than in the past, we would not able to say how much lower with much precision. Why does this matter? Well, the Fed employs an interest rate target. (It does not set the rate.) And under an interest rate targeting regime, as former Chairman Bernanke made clear, the Fed’s “task amounts to using its influence over market interest rates to push those rates toward levels consistent with the equilibrium rate, or—more realistically, its best estimate of the equilibrium rate, which is not directly observable.” The Fed attempts an impossible task when it tries to target an unobservable variable that is constantly in flux. On these grounds, our best course would be to abandon interest rate targeting. We need a better regime. Instead of messing about with interest rates, the Fed should target the level of nominal income. Ideally, it would employ a monetary rule whereby the Fed announces in advance that it will conduct open market operations to maintain a path of nominal income, where the level of nominal income along the path grows at a steady rate of, say, 5% per year. If it fails to hit its target in one period, it will have to make up for its error to return to the path in the next period. Unlike the natural rate of interest, the ideal nominal income path (1) is observable and (2) doesn’t change. I can find no good reason to prefer targeting interest rates over nominal income. As such, the Fed should stop trying to do the impossible. It should, instead, take the more practical approach of targeting nominal income.