Asymmetric Cashless Multipliers: NARDL Evidence on Indonesia

Abstrak

This paper investigates whether positive and negative shocks to payment digitalisation in Indonesia have asymmetric effects on real household consumption and Gross Domestic Product (GDP). Using quarterly data from 2015Q1–2025Q4, a Cashless Payment System Index (CPSI) is constructed through Principal Component Analysis based on debit cards, credit cards, electronic money, and mobile and internet banking transactions. The CPSI is decomposed into positive and negative shocks and estimated using the nonlinear autoregressive distributed lag (NARDL) approach. Twin NARDL models are estimated for real consumption and real GDP while controlling for inflation, interest rates, exchange rates, broad money (M2), banking-sector indicators, and global business-cycle conditions. Bounds tests confirm long-run cointegration in both models. Long-run estimates show that positive cashless-payment shocks increase real consumption and GDP by 0.33% and 0.27%, respectively, while negative shocks affect them by 0.34% and 0.25%. However, Wald tests indicate no significant long-run asymmetry, implying that positive and negative digitalisation shocks have statistically similar effects on economic activity. Robustness checks support this finding. The results suggest that payment digitalisation promotes economic activity in Indonesia, with effects driven primarily by structural factors such as infrastructure expansion rather than behavioural asymmetries.

Kata kunci

• Asymmetric Cointegration
• Cashless Payment System Index
• Household Assumption
• NBER Recession
• Real GDP
• Wald Asymmetry Test