Recently I came across an article on Ft’s Alphaville blog, written by Matthew Klein, arguing about the uselessness of NAIRU models. I had countless similar personal arguments and discussions about the viability of economic models, which usually left unconcluded, leaving both parties disappointed. The typical critique is that one can not reduce the market to two dependent variables and expect a realistic interpretation. I personally believe that if you make this argument you haven’t quite figured out what economic modelling is about. My personal view that you are examining a causality between two variables, leaving everything else out of the correlation (using the much hated Ceteris paribus condition) is not reducing real world, but debunking it. No reasonable modelling economist would argue, that modern countries and societies are built on one or two variables.
As always I missed two parts of mathematical and logical thinking from the post. First, as in any model, here you also have a constant attached to the Phillips function, representing every other part of the system, what may affect the relationship. Mathematically speaking the partial derivative of the real world respective to one variable has to treat the rest of the system constant (ceteris paribus), but does not deny its existence, nor its importance. Logically, this means that any relationship can change, every model can shift due to an external shock, where externality stands for any change coming from the non-examined variables (but not necessarily outside of the economic system). Looking at real life data always produces the net effect of all changes, hence can mask certain relationships, but they exist nonetheless. Relying on just mechanism is just as useless, as denying their existence is ignorance.
As I promised two missing parts of the argument, let me carry on with the less econ-nerdy one. As every part of neo-classical macro modelling, PC also operates with expected inflation and inflation surprise, not merely headline numbers. If the outcome of policy changes will be expected (meaning at least the party of the transaction, who is affected adversely accounts for it) the policy may bear no fruit at all.
It’s not different this time either. Policymakers want to create larger inflation, passing purchasing power from lenders to borrowers, or cover up the losses made during the GFC. The former obviously can’t work if lenders account for higher possible inflation. The second term is more interesting though. Before GFC we, in general, regardless of social status, income, or location, lived in the illusion of wealth… and acted accordingly. I won’t blame anyone, or I could blame everyone, this happens, when group-thinking takes place. The problem is we relied on non-existing wealth, which suddenly disappeared, even faster than it built up. Someone, in real terms, has to foot the bill, and by generating higher than expected inflation the savers can easily found themselves in that undesirable position. And by savers I also mean the want-to-be pensioners.
But this for another time, lets get back to NAIRU for now. As savers are clearly the losers of the inflation surprise, and they tend to be economically more literate, the surprise factor may not materialize. Just think about it, anytime a new QE is announced markets account for the possibility of higher nominal yields, muting the intended long-term effects of the policy. NAIRU is in fact in place, only the smart money is resisting it with every tool it has.