Ekonomista 5 / 2008 p. 577 – 592
The so called neokeynesian Phillips curve is analyzed in the text. The Phillips curve was worked out by Alban William Phillips in 1958 and described in the article: The relationships between unemployment and the rate of change of money wages in the United Kingdom 1861 – 1957. The methodology applied by Phillips was very remarkable for that time. He collected the set of empirical data regarding the economic situation in UK, set it against each other, and concluded that there is a reverse correlation between the unemployment rate and an increase in wages (inflation). The relation suggested that when the inflation was high, the unemployment rate remained low. Phillips formalized the observed relation into the equation. The curve was very interesting and initially seemed to catch one of the important feature of the economy, however it has never been compliant to the relations actually observed in the market. In order to adjust it somehow to the reality and strengthen its predictive power it used to be corrected several times. One of the proposal done by Phelps was even rewarded with Nobel Price. The neokeynesian version is the subsequent step in the curve’s evolution. The most important improvement is that the variables refers to their expected value rather than the actual one. Author analyses few versions of the curve beginning from the Calvo model (this version assumed inflation to be dependent on anticipatory expectation of inflation rate and demand variable) up to the hybrid version (it augments the basic model by the autoregressive mechanism). The most interesting part of the text is however the trial to apply the models to Polish economy and to test it accordingly. The results are dubious or at least ambiguous. According to author’s cautious interpretation the model may not work perfect under condition of imperfect information and disinflation.
Neokeynesizm, mainstream macroeconomics and rational expectation hypothesis seems to be well. Regardless of critiques and attacks from all directions, their apologist do much to maintain the common impression that the economics can really be practiced in that manner, and moreover, that this paradigm is the only one, that can lead us to accurate conclusions. When the subsequent versions of the equation still doesn’t pass the test of correspondence to living market, woe to the market. It looks that Lakatos model of falsification was correct. It is not so easy to falsify the wrong theory by the empirical evidences. Methodologically it is even dubious if the economic (especially neokeynesian) theories are falsifiable (I refer straightly to Milton Freedman’s postulate, that the economic theory should be predictive and therefore falsifiable). The core of the Philipp’s curve theory remains stable and seemingly untouched, although neither of the variables named by Phillips appears in the equation any longer. What remains is only the general idea to bind the variables or its subsequent alterations into function and the strong belief that we are in position sooner or later to worked the function which will generate the accurate market predictions. Maybe it’s time to seriously question that general idea?