New Keynesian Core Model

Evidence and data

Use output gaps, inflation, policy rates, expectations, and price rigidity evidence to discipline the three-equation core.

Calibration or measurement

Price stickiness, policy-rule coefficients, discounting, and intertemporal substitution set the transmission pattern.

Boundaries

• Financial frictions are not automatic.
• Supply chains and sectoral heterogeneity are compressed.
• Expectations assumptions are decisive.

Use guidance

When sufficient

Short-run business-cycle responses to monetary, fiscal, and technology shocks when expectations are forward-looking and nominal rigidities anchor inflation. The model handles the full three-equation core plus investment, habits, and partial indexation when estimated in the Smets-Wouters 2003/2007 form. The right tool when the question requires quantifying impulse responses to identified shocks in a framework with microfounded price and wage setting.

When sketch only

Do not use as the primary tool for questions involving financial-crisis amplification, meaningful household wealth heterogeneity, or the cross-sectional distribution of income. The representative-agent assumption compresses all distributional channels; financial frictions require an additional credit block beyond the basic NK structure.

When to switch

Switch to a financial-accelerator DSGE (dsge:financial-accelerator) when credit spreads and balance-sheet amplification drive the cycle. Switch to HANK (dsge:hank-lite) when the distribution of liquid wealth across households is central to the aggregate demand response. Switch to dsge:soe-nk when the economy is small and open and exchange-rate dynamics are decisive.

Falsification signal

Monetary-shock impulse responses with signs or magnitudes that are systematically inconsistent with the model's predictions across plausible identification schemes falsify the structural parameterization. A persistent price puzzle, where a contractionary shock raises inflation for several quarters before any decline, indicates the policy-rule equation or the information set is misspecified, not minor measurement noise.