Phillips curve Philippic I

This post is a teaser. I am writing that I will be writing about how the Phillips curve is used by the European Commission and in particular EcoFin. I am trying to force myself to write something by promising the web that I will.

Just in general, when reflecting on the other than ideal state of academic macroeconomics, I have comforted myself with the thought that policy makers mostly ignore allegedly micro founded macro and, to the extent they listen to economists at all, use old pretty much Keynesian models roughly based on the work of Keynes Hicks, Solow Samuelson U Friedman and not of Lucas let alone Prescott (yes Friedman would be surprised to be called roughly Keynesian, but, in fact so called New Keynesian economics is largely based on Friedman as noted by Brad DeLong).

I am distressed because I have learned something from Marco Fioramanti about what models policy makers use. The model in question is a highly modified Phillips curve not a silly modern micro-founded macroeconomic model. It is used to estimate the non accelerating wage inflation rate of unemployment (NAWRU not exactly the same as the good old NAIRU because the inflation is wage inflation not price inflation). This is subtracted from unemployment to estimate cyclical unemployment. The estimated Cyclical unemployment is used to estimate potential output and the output gap (with a production function of capital and (employment + cyclical unemployment) and estimated technology (estimating technology is another very tricky issue). The output gap is used to cyclically adjust deficits -- that is to dictate fiscal policy.

The last part is key. At the end of the calculations there aren't recommendations to policy makers. The calculations are part of implementing an international treaty -- the Stability and Growth Pact as modified in 2005.

In order to ensure long-term compliance with the SGP deficit and debt criteria, the member states have since the SGP-reform in March 2005 striven towards achieving their country-specific Medium-Term budgetary Objective (MTO). The MTO is the set limit, that the structural balance relative to GDP needs to equal or be above for each year in the medium-term. Each state select its own MTO, but it needs to equal or be better than a calculated minimum requirement (Minimum MTO) ensuring sustainability of the government accounts throughout the long-term (calculated on basis of both future potential GDP growth, future cost of government debt, and future increases in age-related costs). The structural balance is calculated by the European Commission as the cyclically-adjusted balance minus "one-off measures" (i.e. one-off payments due to reforming a pension scheme). The cyclically-adjusted balance is calculated by adjusting the achieved general government balance (in % of GDP) compared to each years relative economic growth position in the business cycle (referred to as the "output gap"), which is found by subtraction of the achieved GDP growth with the potential GDP growth. So if a year is recorded with average GDP growth in the business cycle (equal to the potential GDP growth rate), the output gap will then be zero, meaning that the "cyclically-adjusted balance" then will be equal to the "government budget balance". In this way, because it is resistant to GDP growth changes, the structural balance is considered to be neutral and comparable across an entire business cycle (including both recession years and "overheated years"), making it perfect to be used consistently as a medium-term budgetary objective.[24][25]

Whenever a country does not reach its MTO, it is required in the subsequent year(s) to implement annual improvements for its structural balance equal to minimum 0.5% of GDP, although it should be noted that several sub-rules (including the "expenditure benchmark") has the potential slightly to alter this requirement.

As a "pact" this can't be unilaterally changed by a national parliament. Here is an extreme (but not misleading) edit ""output gap" ... a country ... is required". The calculated output gap affects what once sovereign countries are required to do.

However, the output gap has to be estimated by a long suffering staff which has to follow clear objective unchanging rules (they not being superior to elected officials as their immediate superiors are). These rules have to make sense and yield sensible results even if the economic model on which they are based doesn't fit recent data well. Since casual discussion of the Stability and Growth pact doesn't clearly distinguish the structural deficit and the ordinary budgetary deficit, large cyclical adjustments would cause extreme suspicion that the rules were being bent.

In particular, that which many countries are required to do depends on a calculation based on an accelerationist Phillips curve. Behind the calculation there is a model in which it is assumed that unemployment affects the change in the rate of wage inflation and in which a decrease from 11% to 9% is just like a decrease from 1% to -1%. This means that, to the extent that the Stability and Growth Pact is involved, policy is absolutely not based on consideration of Phillips curves with anchored expectations.

The reassuring claims about policy and Phillips curves made to Paul Krugman -- " the usual response from model-oriented public officials and research staff at policy institutions is to say that what they work with now is a Phillips curve with “anchored” expectations " -- are to this important extent, false.

Krugman calculated a Euro area output gap using an anchored Phillips curve. Even he is astounded (and unconvinced) by his calculated output gap of 8%. Using such a gap for the correction would allow a fundamental transformation of fiscal policy in many Euro area countries. These do not including those, such as Germany, where the pact isn't binding, or Greece which can't borrow even if it had EcoFin permission, because no one will lend to them. However, they most definitely include Italy.

I don't take Krugman's calculation more seriously than he does (another post) but it is easy to understand why the binding authoritative calculations with the force of an international treaty are completely different. One reason is that they are based on an accelarationist Phillips curve estimates of which imply double digit natural rates of unemployment (which have shot up since 2008). The other (and probably more important) explanation is that such a conclusion is unacceptable to the Eurocrats who claim to be technicians not dictators, but who accept no limits on their power and will never accept markedly higher deficits.

Still the demonstrable fact that the official and binding calculations are based on a model which is wildly at odds with the data might have some impact on someone somewhere sometime somehow.