Least Squares to get slope of Phillips curve
Posted: Tue Nov 22, 2011 5:11 am
Dear all
I am about to estimate Brazilial NAIRU using Ball-Mankiw approach. This approach is based on the following
standard expectations-augmented specification for the Phillips curve:
π(t) = π(t-1) − a [U(t) − NAIRU ] + vt
where π(t) is the actual rate of inflation, π(t-1) the expected rate of inflation under adaptive expectation, U(t) the actual
rate of unemployment, and vt a supply shock. This is rewritten as:
U(t) + (delta_π / a) = NAIRU + (vt / a)
For a given value of the slope of the Phillips curve (a) obtained using LS, the left-hand side of the
above expression can be computed from the data, yielding an estimate of NAIRU + (vt / a), where NAIRU can be extracted using a HP Filter.
I HAVE 2 DOUBTS:
1) I have tried to get value of Phillips curve slope (a) using LS, but I do not know how to estimate it considering I dont have the NAIRU data yet.
2) After obtaining (a) and getting the vallues of NAIRU + (vt / a), which are the staps to use HP Filter?
THANK YOU VERY MUCH
I am about to estimate Brazilial NAIRU using Ball-Mankiw approach. This approach is based on the following
standard expectations-augmented specification for the Phillips curve:
π(t) = π(t-1) − a [U(t) − NAIRU ] + vt
where π(t) is the actual rate of inflation, π(t-1) the expected rate of inflation under adaptive expectation, U(t) the actual
rate of unemployment, and vt a supply shock. This is rewritten as:
U(t) + (delta_π / a) = NAIRU + (vt / a)
For a given value of the slope of the Phillips curve (a) obtained using LS, the left-hand side of the
above expression can be computed from the data, yielding an estimate of NAIRU + (vt / a), where NAIRU can be extracted using a HP Filter.
I HAVE 2 DOUBTS:
1) I have tried to get value of Phillips curve slope (a) using LS, but I do not know how to estimate it considering I dont have the NAIRU data yet.
2) After obtaining (a) and getting the vallues of NAIRU + (vt / a), which are the staps to use HP Filter?
THANK YOU VERY MUCH