<![CDATA[This study investigates the impact of population ageing and elderly labour on economic growth in Jawa Timur, an early-ageing province in Indonesia. The analysis utilises a balanced panel of 38 regencies from 2017 to 2024, employing one-way fixed-effects and two-stage least squares (2SLS) estimation. To address potential endogeneity, lagged infant mortality and lagged access to safe drinking water serve as instruments for the elderly dependency ratio. Initial fixed-effects estimates indicate a demographic burden. However, after correcting for endogeneity, the 2SLS results reveal no evidence of a causal effect of the elderly dependency ratio, elderly labour force participation, or life expectancy on economic growth. Thus, a higher proportion of elderly individuals does not causally reduce economic growth. Conversely, elderly educational attainment demonstrates a positive and statistically significant causal effect on growth, a result that remains robust after controlling for poverty. As elderly individuals completed their formal education several decades ago, this finding underscores the long-term benefits of historical investments in education. The results support the third demographic dividend hypothesis in a middle-income regional context, with the principal transmission channel operating through human capital rather than labour supply. Policy implications are as follows: in the short term, labour metrics should be reoriented, and data quality improved; in the medium term, targeted adult education programmes should be reassessed; and in the long term, current educational investments will shape the productivity of the elderly in the 2080s.]]>
