"The Phillips Curve model is wrong because Keynesians get causality reversed."

"In fact, the Keynesian Phillips curve model is simply wrong. It’s not wrong because there is no relationship between inflation and unemployment. A sharp fall in both wage and price inflation tends to be associated with a temporary rise in unemployment. Rather the Phillips Curve model is wrong because Keynesians get causality reversed. They assume that causation goes from economic overheating to wage and price inflation, whereas the opposite is more nearly true. To be precise, it is unexpected increases in nominal growth in spending that cause both rising inflation and falling unemployment."

~ Scott Sumner, from his post 'Fortunately, this isn't the Volcker disinflation'