Hauser: Nonlinear Phillips curve shapes RBA policy strategy

Phillips' original insight into the nonlinear nature of the Phillips curve was largely neglected for decades in academic and policy circles, favoring a linear assumption.

This proved a 'costly mistake' when, post-Covid, global inflation surged amid near-full employment, exactly what a nonlinear curve would predict.

This event sparked renewed interest in providing robust micro-foundations for nonlinearity.

Australia, and the RBA, had not forgotten this; seminal work by Debelle and Vickery showed a nonlinear model fitted Australian data better, leading the RBA to adapt its forecasting suite.

Yet, even these models underestimated the post-Covid inflation pickup, partly due to difficulties in identifying underlying impulses and fully capturing the nature of the nonlinearities.

Models were often estimated over periods on the flatter part of the curve.

This history underscores that nonlinearity has first-order implications for monetary policymakers, demanding detailed understanding of its steepness, position, and shifts.

Recent research highlights three key sources of Phillips curve nonlinearity.

Downward nominal wage rigidity means wages rarely fall; in weak markets, further weakening has little effect, but in tight markets, wages grow strongly.

The highly nonlinear Beveridge curve, describing vacancies to unemployment, is crucial: high vacancies at low unemployment compel firms to offer higher wages.

Nonlinear hiring costs also contribute, accelerating quickly in tight labour markets due to increased effort and recruiter capacity constraints.

While not affecting the curve's shape, inflation expectations can shift it up or down, perpetuating inflationary shocks.