Abstract:
We implement a quantitative empirical test of the fiscal theory of the price level (FTPL) model via indirect inference, comparing it to a standard New Keynesian model. The FTPL alternative creates a serious instability problem because it triggers a 'doom loop'in which inflation pushes up interest rates which in turn pushes up deficits and debt and so inflation. Without some sort of endogenous feedback response this instability prevents the model from solving in finite space; a fiscal 'whatever it takes' commitment to prevent unstable paths fails to create stablity because no steady state inflation solution exists. This is the case whether we embed FTPL in an otherwise conventional New Keynesian model or in a classical RBC model. We then went on to look for endogenous responses of government spending and tax to the economy - a Fiscal Rule - which might render the FTPL model suffi ciently stable to be testable. We found such a Rule: in it spending stabilises the output gap while tax responds to inflation, with an inflation cap ('tax reform') - such that if inflation exceeds some high rate it overrides the FTPL terminal condition by inserting whatever terminal surplus will cap inflation at this rate. With this rule in place we can solve an RBC version of the model without triggering intolerable volatility; this version is on the test rejection borderline on our full postwar sample whereas the standard New Keynesian model passes the test easily.