The missing equation
Scott Sumner has written dozens of blog posts trying to explain to why the fiscal multiplier is zero if the central bank targets the NGDP level, the price level or inflation. Said in another way Scott – as do I – strongly believe that the impact of fiscal policy strongly dependent the monetary policy reaction to fiscal tightening or fiscal easing (Even today Scott has a discussion of the fiscal multiplier). In fact I don’t even think it is meaningful to talk about fiscal policy as something that can “stimulate” demand. Hence, in a pure barter economy we cannot imagine fiscal policy having any impact on demand as demand always will equal supply in a barter economy. The famous Say’s Law holds in a barter economy and as such there would be full crowding out of fiscal policy. Hence, fiscal policy will only have an impact on demand if monetary policy “plays along”. Our view is however far from the consensus among economists. Rather most economists think that you can use fiscal policy to “manage” nominal spending/demand. Even economists who in general do not find activist fiscal policy desirable tend to think that fiscal policy can impact nominal demand. Today after working on some macroeconomic models myself I finally realised that the problem is that the “models” that most economists have in their heads are missing an equation (or at least one equation). Hence, most economists – and here I am talking about practicing macroeconomists like central bank economists or financial sector economists like myself – tend to give very little or no attention at all to the monetary policy regime of the economy they are analysing. Therefore, the missing equation in most “models” is the policy reaction function of the central bank. And it might even be worse as it seem like most practical macroeconomists tend to assume that the central bank “accommodates” any change in fiscal policy so when fiscal policy is eased the central bank plays along and just automatically increase the money supply so to increase nominal GDP (with sticky prices that will increase RGDP and reduce unemployment). In such a world the “fiscal” multiplier obviously is positive. The problem is, however, that it is not really fiscal policy as such which is increasing NGDP, but the central bank’s “automatic” easing of monetary policy. This assumption clearly is counterfactual. Anybody who has watched the actions of for example the ECB over the last couple of years would know that central banks certainly does not automatically play along. The fact that many economists do not realise that they are missing an equation in their (mental) models also means that they completely fail to realise that the causality in their models are hugely dependent on the monetary policy reaction function. This is also why it is so hard for many to comprehend that monetary policy can work with long and variable leads (See my discussion of monetary policy leads and lags here – the discussion is basically a variation of the discussion in this post). If you don’t believe me then try to have a look at the regular macroeconomic reports of central banks around world. In most of these reports the story of causality is pretty much a simple national account model, where increases in private consumption, investment and government spending etc. are described as “automatically” leading to an increase in real and nominal GDP. The monetary policy reaction is given little or no attention. This description is of course not totally fair as many central banks are using so-called DSGE models that take explicitly take into account (some kind of) the monetary policy reaction. However, one thing is a model for simulations – another thing is the verbal description of the economy. See for example here: “The slowing domestic demand growth observed since 2010 Q4 turned into a noticeable annual decline in 2011 Q3. This decline was due to all its components, but most of all to change in inventories…Household and government consumption, whose annual decline intensified, affected domestic demand to a lesser extent. A renewed decline in fixed investment also contributed to the contraction in domestic demand.” This is from the Czech central bank’s latest quarterly inflation report (page 35). We are told that demand slowed because of weak private consumption and tighter fiscal policy have lowered demand growth. However, what role did monetary policy play in this? Isn’t nominal demand sorely determined by monetary policy? So we cannot see the slowing of Czech demand without taking into account monetary policy. In fact it is interesting that when central bankers describe the ups and downs in the economy nearly never hold themselves accountable. If inflation overshoots the inflation target we rarely (in fact never) hear central banks say “the failure to fulfil our inflation target was due to our overly loose monetary policy”. I wouldn’t really expect that and frankly I also hate admitting being mistaken. But this is nonetheless telling of the general tendency for macroeconomists – including those working for central banks - to fail to realise the importance of the monetary policy reaction function. I should of course stress that I am certainly no saint. I am as lazy as any other economist and I most admit to many times both in written and spoken form to have argued along the lines of the “national account model”, but I hope that I at least know when I am intellectually lazy. PS The mentioning of the Czech central bank’s inflation report above is not meant as a specific critique of my friends at the CNB. The economists at the CNB are certainly capable economists and it should be noted that the DSGE model used by the CNB’s research department explicitly tries to take into account the monetary policy reaction function of the board of the CNB.