Microfoundations As Well As The Phillips Curve

This is a rather long shipping service that follows from my before one on Faustian bargains. Mainly for economists.

Although the debate close microfoundations unremarkably talks close models, it is possible inward about cases to also verbalise close private relationships. This has the reward of simplicity. I intend the Phillips bend is a overnice illustration of many of the primal issues inward the debate - fifty-fifty for those who produce non believe inward it! (I brand the to a greater extent than full general declaration here.)

First, it illustrates for me why the debate is not, I repeat not, close whether the microfoundations approach is useful. Of course of didactics it is. Let’s intend close the Phillips bend before microfoundations. This traditional Phillips bend had inflation at fourth dimension t depending on expectations of inflation also at fourth dimension t, as well as the output gap. This seemed at the fourth dimension plausible: y'all could enjoin a storey close lags inward wage setting, so actors inward the labour marketplace had to approximate the relevant existent wage etc.

This is an instance of the dangers of promotion hoc theorising. The lags involved inward wage decisions arise primarily because of contracts: reward are beingness fixed non only for the adjacent period, but for a number of periods. Once y'all produce the microfoundations properly, y'all observe that inflation today depends on expected inflation next period: the New Keynesian Phillips bend (NKPC).

A detail? Unfortunately the promotion hoc version implied that the output gap would movement amongst expectation errors. So the dynamics of the concern wheel became all close the inflation expectation fault process. This was ever implausible, fifty-fifty if y'all did non believe inward rational expectations. With rational expectations, it appeared to deny the existence of Keynesian concern cycles, as well as led to a whole generation of students beingness taught that anticipated monetary policy was ineffective.

So the traditional Phillips bend illustrates the dangers of ‘ad hoc’ theorising. But the relevant debate is non close whether microfounded relationships are useful, but whether they should hold out the only game inward the academic town. So at nowadays allow me give y'all diverse reasons why they should not, based on the Phillips bend example.

First, although diverse microfounded models propose inflation should depend on expected inflation adjacent period, the empirical bear witness is to a greater extent than equivocal. Influenza A virus subtype H5N1 number of studies receive got found that inflation today depends on both expected inflation adjacent period, but also on actual inflation concluding menses (‘inflation inertia’). There are also empirical regularities (e.g. Phillips bend loops) that are hard to rationalise amongst only a NKPC. As yet in that location is no to a greater extent than oftentimes than non accepted microfoundation for why this inflation inertia mightiness occur, but given the primal utilization this equation plays inward policy it is of import to facial expression at the consequences of inflation inertia. That is what Steinsson did inward a 2003 newspaper inward JME, for example. [1] Now Steinsson’s introduction of ‘rule of thumb’ cost setters was alone promotion hoc, amongst no microfoundation (traditional or behavioural) offered. Would those who decry theory past times ‘hand waving’ equally promotion hoc intend that article should non receive got been published inward the JME?

At this indicate about mightiness say that this instance illustrates that in that location is no occupation - papers amongst non-microfounded relationships produce larn published. Yet everyone knows that in that location is a large effective bias against publishing models that are non fully microfounded: Steinsson’s newspaper is the exception rather than the rule. (See my discussion of middlebrow models.) Indeed I started seriously thinking close these issues afterwards I attended a conference where ane respected player attacked a couplet of papers that did include inflation inertia, proverb nosotros must ‘respect microfoundations’. It was this that gave rising to my distinction betwixt microfoundations purists as well as microfoundations pragmatists.

This instance illustrates that nosotros are oftentimes non confronted amongst a selection betwixt ii relationships: ane microfounded as well as ane not. Instead nosotros receive got the promotion hoc relationship, as well as nosotros are withal waiting for the microfounded version. Microfoundations takes time. There is a uncomplicated argue for this: microfoundations modelling is difficult, as well as it oftentimes relies on sure enough tricks to retain tractability. Those tricks may non hold out at in ane lawsuit obvious, as well as may convey fourth dimension to larn ‘acceptable’. So inward the early on 1970s, before New Keynesian theory was developed, nosotros only had the traditional (ad hoc) Phillips curve. In those circumstances, are nosotros seriously suggesting that academic macroeconomics should receive got ignored this human relationship because it was promotion hoc? Surely nosotros desire to model what we tin shipping away see, as well as non only what nosotros tin shipping away microfound.

Second, an aggregate human relationship tin shipping away hold out much to a greater extent than robust than whatsoever exceptional microfoundation. You larn the resultant that inflation depends on expected inflation adjacent menses from dissimilar models of cost rigidity: Calvo contracts of course of didactics (which tin shipping away hold out said to mimic the bear on of bill of fare costs - encounter below), but also fixed contracts. Why necktie things downwardly to ane exceptional cost rigidity story, rather than going straight to an aggregate human relationship which is compatible amongst to a greater extent than than ane microfoundation?

The response is oftentimes consistency. You necessitate to demonstrate that your microfoundations are consistent across the model equally a whole. Now this is clearly truthful for something similar consumption as well as labour furnish decisions. However models of nominal inertia are pretty separable from the remainder of a model, or nosotros employ tricks to ensure that they are. So the reasons for having to piece of work through the Calvo set-up every fourth dimension anyone wants to utilization a NKPC hardly seem compelling.

Third, only how neat are these microfoundations anyway? The Calvo contract storey has the expected probability of changing prices exogenous to the firm, as well as it is non obvious why this is a reasonable thing to assume. In practise microfoundation modellers moving ridge their hands equally much equally anyone, talking close the bear on of bill of fare costs without formally modelling them. As I debate here, this is an ‘as if’ shape of justification which is an of import footstep removed from establishing model consistency based on only deep parameters.

If all this were non compelling enough, a belief that the microfoundations methodology is the only agency of ‘doing macroeconomics properly’ has unfortunate side effects. It moves modellers away from the existent basis as well as into an imaginary basis of yeoman farmers as well as the like. So about volition react to the concluding paragraph past times agreeing, proverb that they were ever dubious that Calvo contracts represented proper microfoundations, as well as that it is amend thence to ignore cost stickiness until a to a greater extent than acceptable microfoundation tin shipping away hold out found.

This is a bizarre reaction. The number amongst cost rigidity is that in that location are also many potential microfoundations, non also few. (See this good known newspaper past times Alan Blinder for example.) The occupation is inward reality ii fold. First, each argue is quite complex, so they are hard to easily contain inward a model. Tractability is the problem, non evidence. Second, fifty-fifty if y'all tin shipping away produce this for ane exceptional motivation for cost stickiness (e.g. Calvo contracts for bill of fare costs), y'all are ignoring all the others - y'all ignore heterogeneity. Is this because y'all produce non intend these other reasons matter, or is it because the model is complicated plenty equally it is? The latter, obviously. Yes, of course of didactics it is of import to produce microfoundations modelling amongst heterogeneity - but this oftentimes results inward a loss of clarity. If this heterogeneity inward practise leads to a really similar aggregate outcome, why non piece of work straight to the aggregate?

We are inward danger of fooling ourselves here. We pick out Calvo contracts to microfound cost rigidity non because nosotros believe this is how the bulk of firms behave, but because it is tractable, as well as gives us an aggregate human relationship (the New Keynesian Phillips curve) that is consistent amongst the data. Yet nosotros as well as then hold to debate that if whatsoever other aspect of the model is internally inconsistent inward terms of its microeconomics amongst Calvo contracts, the model should hold out rejected. Finally, nosotros as well as then claim that anyone who uses a New Keynesian Phillips bend without doing this microfoundations banking concern jibe is non doing serious macroeconomics.

So nosotros pick out a microfoundation because it gives us the aggregate response suggested past times the data, as well as non because of bear witness that this microfoundation is appropriate. We as well as then insist that everything inward that model has to hold out consistent amongst this microfoundation, as well as that our model has been built from only thinking close what private agents do. Have nosotros only replaced promotion hoc amongst shipping service hoc?

So, to summarise. The Phillips bend shows why microfoundations macro is actually useful. However it also shows why microfoundations modelling should non hold out the only agency of doing proper macroeconomics.

[1] Steinsson, J (2003), Optimal Monetary Policy inward an economic scheme amongst Inflation Persistence, Journal of Monetary Economics, 50, 1425-1456