RDP 8909: Optimal Wage Indexation, Monetary Policy and the Exchange Rate Regime Appendix 2: Testing the Overidentifying Restrictions
December 1989
- Download the Paper 753KB
The overidentifying restrictions were tested by constructing the following statistic for each estimated equation:
where T is the number of observations and R2 is the coefficient of determination from a regression of the IV residuals on the available instruments. Q is distributed as Chi-square with degrees of freedom equal to the number of overidentifying restrictions. Its marginal significance level, for each test, is reported in Table A2.1.
| COUNTRY | EQUATION | |||
|---|---|---|---|---|
| wage | money supply | aggregate demand | price | |
| Australia | 0.540 | 0.423 | 0.000 | – |
| Austria | 0.714 | 0.614 | 0.000 | – |
| Canada | 0.775 | 0.127 | 0.000 | – |
| France | 0.175 | 0.806 | 0.000 | – |
| Germany | 0.012 | 0.901 | 0.000 | – |
| Italy | 0.344 | 0.009 | 0.000 | – |
| Japan | 0.199 | 0.509 | 0.000 | – |
| UK | 0.034 | 0.162 | 0.000 | – |
| USA | 0.424 | 0.845 | 0.000 | 0.000 |
The test results show that the overidentifying restrictions in the money supply and wage equations are not rejected at conventional significance levels. However, the evidence points to some correlation between the shocks to aggregate demand and the other shocks.