Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This study investigates the impact of uncertainty shocks on output in New Zealand between 1985Q2 and 2018Q4 using the Internal and External Instrument versions of the Structural Vector Autoregression model. While the existing literature has relied almost exclusively on the External Instrument approach, this method requires the invertibility of the uncertainty shock, whereas the Internal Instrument approach does not. We formally test for, and reject, invertibility in our application, reflecting that uncertainty frequently arises from concerns about future policy or economic fundamentals. As the two instrumental variable models produce quantitatively different impulse responses for output, we empirically illustrate the importance of testing for invertibility and using the results to guide model selection. We also find that the effects of uncertainty shocks on New Zealand’s output are larger than previously documented — with investment a key transmission channel — suggesting that counter-cyclical policy may need to respond more aggressively to uncertainty-driven downturns.