Econ 398 of October 21, 2015
• Friedman presidential address (Dec 1967)
» cf. Krugman’s obituary of Friedman
» cf. Friedman’s Nobel laureate address (1976)
– context: inflation had not been an issue
– monetary policy had not been a core component of macroeconomic policy
» exceptions included Yale’s James Tobin, though Tobin did not agree with Friedman on politics
– context: Friedman & Schwartz Monetary History claim that the Fed worsened the Great Depression [Friedman in his public persona transformed this into the Fed caused the GD]
= claim that there is a natural rate of unemployment and that this meant money was neutral in the long run: only real variables matter
› but the Fed could shift the economy in the short run, for better and (what he emphasized) for worse
» hence rules not discretion!!
= claim that stable policy was thus ideal
– not really disputed in principle by anyone, but how big are the gains?
• technical: NAIRU and the inflation-augmented Phillips Curve
= clearer in his Nobel lecture in that his Presidential Address did not include graphs
∑ overall this represented a major challenge to accepted theory and particularly policy
= [did not discuss] it proved timely, because it then provided an explanation for the poor economic performance of the 1970s, which saw both (for the US) high inflation and (for the postwar era) high unemployment
• Taylor Rule
it = πt + r*t + aπ (πt – π*t) + ay (yt – ȳt) [not the notation used in class]
– Taylor used aπ = ay = ½
– monetary policy should thus reflect both booms/busts [in practice ȳ is estimated using some version of NAIRU or a Phillips Curve, or by projecting past growth either through a moving average or a simply regression]
– if y is 1 percentage point too high, then raise i by ½ percentage point
– if π is 1 percentage point too high, then raise i by 1½ percentage points
» [regroup the π terms to get [1+ ½]π + r* – ½π*
» Taylor set target inflation at π* = 2 and r* = 2 so that at π = 2 we get i = 4.
= assumes we can usefully estimate ȳ and r*
– ȳ is NAIRU, the level of GDP consistent with stable inflation under a long-run Phillips Curve
– r* is the natural rate of interest, reflecting the growth potential of the economy
= assumes we have good real-time measures of inflation and GDP
= does not include an exception for financial stability exception, though acknowledged in Taylor’s discussion so could in principle be tacked on as “α“
› [did not discuss!] in practice the Fed does not typically change its target interest rate by large increments, so this should be rewritten to reflect a “smoothing” of rates
› [did not discuss!] implicitly assumes a “normal” economy with positive inflation and a positive implied policy i
› [did not discuss!] what weights to employment versus inflation? no particular objective reason to use ½ (though Taylor does show that it seems to be the implicit rule used by the Fed)
› [did not discuss!] unless the Fed is explicit in real time about what it believes is happening to actual employment & inflation it is not transparent in practice.
» implicit in this is that monetary policy is not immediate in its macroeconomic impact, so the Fed needs to be making decisions on what it believes will be the state of the economy 4-6 quarters hence.
» transparency thus requires that the Fed publicize its projections