Editing New Keynesian economics (section)

===1990s===

====New neoclassical synthesis====

In the early 1990s, economists began to combine the elements of new Keynesian economics developed in the 1980s and earlier with [[Real Business Cycle Theory]]. RBC models were dynamic but assumed perfect competition; new Keynesian models were primarily static but based on imperfect competition. The [[new neoclassical synthesis]] essentially combined the dynamic aspects of RBC with imperfect competition and nominal rigidities of new Keynesian models. Tack Yun was one of the first to do this, in a model that used the [[Calvo (staggered) contracts|Calvo pricing]] model.<ref>Yun, Tack (April 1996). "Nominal price rigidity, money supply endogeneity, and business cycles". ''Journal of Monetary Economics'' 37(2–3). Elsevier. pp. 345–370.</ref> Goodfriend and King proposed a list of four elements that are central to the new synthesis: intertemporal optimization, rational expectations, imperfect competition, and costly price adjustment (menu costs).<ref>Goodfriend, Marvin; King, Robert G (1997). "The New Neoclassical Synthesis and the Role of Monetary Policy". ''NBER Macroeconomics Annual''. NBER Chapters (National Bureau of Economic Research) 12: 231–83, {{JSTOR|3585232}}.</ref><ref>Snowden and Vane 2005, p. 411</ref> Goodfriend and King also find that the consensus models produce certain policy implications: whilst monetary policy can affect real output in the short-run, but there is no long-run trade-off: money is not [[neutrality of money|neutral]] in the short-run but it is in the long-run. Inflation has negative welfare effects. It is important for central banks to maintain credibility through rules based policy like inflation targeting.

====Taylor Rule====

In 1993,<ref>{{cite journal |last=Taylor |first=John B. |year=1993 |title=Discretion versus Policy Rules in Practice |journal=Carnegie-Rochester Conference Series on Public Policy |volume=39 |url=http://www.stanford.edu/~johntayl/Papers/Discretion.PDF |pages=195–214 |doi=10.1016/0167-2231(93)90009-L }} (The rule is introduced on page 202.)</ref> John B Taylor formulated the idea of a '''[[Taylor rule]]''', which is a reduced form approximation of the responsiveness of the [[nominal interest rate]], as set by the [[central bank]], to changes in inflation, [[Gross domestic product|output]], or other economic conditions. In particular, the rule describes how, for each one-percent increase in inflation, the central bank tends to raise the nominal interest rate by more than one percentage point. This aspect of the rule is often called the '''Taylor principle'''. Although such rules provide concise, descriptive proxies for central bank policy, they are not, in practice, explicitly proscriptively considered by central banks when setting nominal rates.

Taylor's original version of the rule describes how the nominal interest rate responds to divergences of actual inflation rates from ''target'' inflation rates and of actual gross domestic product (GDP) from ''potential'' GDP:

<math display="block">i_t = \pi_t + r_t^* + a_\pi  ( \pi_t - \pi_t^* )  + a_y ( y_t - y_t^* ).</math>

In this equation, <math>\,i_t\,</math> is the target short-term [[nominal interest rate]] (e.g. the [[federal funds rate]] in the US, the [[Official bank rate|Bank of England base rate]] in the UK), <math>\,\pi_t\,</math> is the rate of inflation as measured by the [[GDP deflator]], <math>\pi^*_t</math> is the desired rate of inflation, <math>r_t^*</math> is the assumed equilibrium real interest rate, <math>\,y_t\,</math> is the logarithm of real GDP, and <math>y_t^*</math> is the logarithm of [[potential output]], as determined by a linear trend.

====New Keynesian Phillips curve====

The New Keynesian Phillips curve was originally derived by Roberts in 1995,<ref>{{cite journal |last=Roberts |first=John M. |year=1995 |title=New Keynesian Economics and the Phillips Curve |journal=[[Journal of Money, Credit and Banking]] |volume=27 |issue=4 |pages=975–984 |jstor=2077783 |doi=10.2307/2077783}}</ref> and has since been used in most state-of-the-art New Keynesian DSGE models.<ref>{{cite book |last=Romer |first=David |year=2012 |chapter=Dynamic Stochastic General Equilibrium Models of Fluctuation |title=Advanced Macroeconomics |publisher=McGraw-Hill Irwin |location=New York |pages=312–364 |isbn=978-0-07-351137-5 }}</ref> The new Keynesian Phillips curve says that this period's inflation depends on current output and the expectations of next period's inflation. The curve is derived from the dynamic Calvo model of pricing and in mathematical terms is:

<math display="block">\pi_{t} = \beta E_{t}[\pi_{t+1}] + \kappa y_{t}</math>

The current period {{mvar|t}} expectations of next period's inflation are incorporated as <math>\beta E_{t}[\pi_{t+1}]</math>, where <math>\beta</math> is the discount factor. The constant <math>\kappa</math> captures the response of inflation to output, and is largely determined by the probability of changing price in any period, which is <math>h</math>:
 
<math display="block">\kappa = \frac{h[1-(1-h)\beta]}{1-h}\gamma</math>.

The less rigid nominal prices are (the higher is <math>h</math>), the greater the effect of output on current inflation.

====Science of monetary policy====

The ideas developed in the 1990s were put together to develop the new Keynesian [[dynamic stochastic general equilibrium]] used to analyze monetary policy. This culminated in the three-equation new Keynesian model found in the survey by [[Richard Clarida]], [[Jordi Gali]], and [[Mark Gertler (economist)|Mark Gertler]] in the ''[[Journal of Economic Literature]]''.<ref>Clarida, Galí, and Gertler (2000)</ref><ref>{{cite journal |last1=Clarida |first1=Richard |last2=Galí |first2=Jordi |last3=Gertler |first3=Mark |year=2000 |title=Monetary Policy Rules and Macroeconomic Stability: Evidence and Some Theory |journal=[[The Quarterly Journal of Economics]] |volume=115 |issue=1 |pages=147–180 |doi=10.1162/003355300554692 |citeseerx=10.1.1.111.7984 |s2cid=5448436 }}</ref> It combines the two equations of the new Keynesian Phillips curve and the Taylor rule with the ''dynamic IS curve'' derived from the optimal [[Random walk model of consumption|dynamic consumption equation]] (household's Euler equation).

<math display="block">y_{t}=E_{t} y_{t+1} - \frac{1}{\sigma}(i_{t} - E_{t}\pi_{t+1})+v_{t}</math>

These three equations formed a relatively simple model which could be used for the theoretical analysis of policy issues. However, the model was oversimplified in some respects (for example, there is no capital or investment). Also, it does not perform well empirically.