Time-Varying Monetary–Fiscal Policy Interaction and Inflation Dynamics: Evidence from Indonesia using a TVP-VAR-SV Approach

Technical Summary: Time-Varying Monetary–Fiscal Policy Interaction and Inflation Dynamics: Evidence from Indonesia using a TVP-VAR-SV Approach

1. Problem Statement and Motivation
The central problem addressed is the limited understanding of how monetary and fiscal policies jointly shape inflation in Indonesia, particularly given the country's history of regime shifts, global crises, and post-pandemic fiscal expansion. While the Fiscal Theory of the Price Level (FTPL) suggests that inflation dynamics may be driven by fiscal expectations when the fiscal authority behaves "actively," empirical literature for Indonesia has relied on fixed-parameter Structural VAR (SVAR) models. These static models fail to capture the time-varying nature of policy transmission, which is critical in an economy characterized by structural breaks (e.g., the 2005 adoption of inflation targeting) and asymmetric responses to external shocks. The authors argue that averaging parameters across distinct regimes risks obscuring the very dynamics—such as fiscal dominance or the erosion of monetary policy space—that are most relevant for policy coordination between Bank Indonesia and the Ministry of Finance.

2. Methodology
The study employs a Time-Varying Parameter Vector Autoregression with Stochastic Volatility (TVP-VAR-SV) framework, following the specification of Primiceri (2005). This approach is selected to endogenously estimate structural changes without imposing exogenous break points.

• Data: The analysis utilizes quarterly data from 2000Q1 to 2025Q1 (101 observations).
• Variables: The system includes five variables: Exchange Rate (EXR), Government Spending (GS), Industrial Production Index (IPI), Consumer Price Index (CPI), and the BI Rate (BIR).
• Identification: A recursive Cholesky decomposition is applied with the ordering: EXR → GS → IPI → CPI → BI Rate. This ordering assumes the exchange rate is most exogenous within the quarter, followed by fiscal impulses, real activity, prices, and finally the monetary policy instrument.
• Estimation: The model is estimated using Bayesian Gibbs sampling. Priors include a Minnesota prior for initial coefficients and Inverse-Wishart/Inverse-Gamma distributions for innovation covariances and volatilities. Convergence is verified via Geweke and Heidelberger–Welch tests, with Effective Sample Sizes (ESS) ranging from 9,606 to 10,000.
• Validation: A Sup-Wald test confirms significant parameter instability (F = 2,604.59 at the 2005Q4 breakpoint), justifying the time-varying approach over a static VAR.

3. Key Contributions
The paper makes four specific contributions to the literature:

• Methodological: It is one of the first studies to apply the TVP-VAR-SV framework specifically to analyze monetary–fiscal interactions in Indonesia within an FTPL context, integrating time-varying parameters, domestic fiscal regimes, and external spillovers in a unified model.
• Theoretical: It provides an empirical test of FTPL predictions in a small open developing economy, examining whether fiscal shocks drive inflation more than monetary shocks during periods of fiscal expansion.
• Empirical: It extends the temporal coverage of existing Indonesian studies (which typically end in 2019 or 2022) to include the 2022–2025 post-pandemic normalization period, a phase previously unanalyzed under a time-varying framework.
• Policy: It offers an empirical basis for evaluating the coordination mechanisms between Bank Indonesia and the Ministry of Finance, specifically addressing the "burden-sharing" and subsidy dynamics observed in recent years.

4. Key Results
The analysis yields three principal findings regarding inflation dynamics and policy interaction:

• Negligible Monetary Impact: The contribution of BI Rate shocks to the forecast error variance of inflation is negligible across all horizons and episodes. This is consistent with a "passive" monetary regime where the central bank's independent ability to control prices is constrained.
• Fiscal and External Dominance: Government expenditure accounts for between 3.0% and 27.8% of CPI variance depending on the episode, while exchange rate shocks play a larger role in certain periods (reaching 26.4% in 2001Q3). The Historical Decomposition confirms that fiscal and external channels have historically contributed more to inflation dynamics than pure monetary shocks.
• Erosion of Monetary Space: A critical finding is the rising share of BI Rate variance attributable to government expenditure shocks, which increased from 14.2% in 2008Q4 to 60.2% in 2022Q4. This indicates that the post-pandemic fiscal expansion has progressively narrowed the effective space for independent monetary policy.
• Price Puzzle and Anomalies: The study observes a "price puzzle" (positive short-run CPI response to rate hikes) and a specific anomaly in 2016Q3 where rate hikes coincided with persistent inflation, interpreted as a signal that monetary policy could not offset expansionary fiscal expectations.

5. Significance and Claims
The authors claim that their findings are consistent with the predictions of the Fiscal Theory of the Price Level regarding an active fiscal regime. The study concludes that the effectiveness of monetary policy in Indonesia is highly conditional on the prevailing fiscal stance. Specifically, the paper argues that effective post-COVID monetary normalization in Indonesia cannot be achieved through unilateral interest rate adjustments alone; rather, it requires coordinated fiscal consolidation. The results suggest that the "harmonization" observed in previous fixed-parameter studies may actually reflect a structural shift toward fiscal dominance, where monetary policy increasingly accommodates fiscal conditions rather than operating with symmetric independence.

The paper maintains a modest stance, acknowledging that the identification relies on Cholesky ordering and that the TVP-VAR-SV evaluates dynamic patterns consistent with FTPL rather than proving the intertemporal mechanism in a fully structural causal sense. However, the robustness checks (including alternative variable orderings and lag sensitivities) confirm that the core finding—fiscal dominance over monetary transmission in recent episodes—is stable.