Economic Forecasting for Climate-Responsive Policy
We rarely question a doctor's diagnosis, especially from someone experienced, even without further explanation. That trust grows from a belief that long training sharpens intuition. We seldom extend the same trust to economic projections, whose usefulness is often doubted. Yet like doctors, economists develop intuition through long experience.
It is true that economic forecasts often miss. Daniel Kahneman, a Nobel laureate in economics, noted that economic predictions are frequently even less accurate than random guesses. One study found they were correct only about 23% of the time, even though 53% of forecasters were confident they were right (Campbell and Moore, 2024). The economic system may not be as complex as the human body, but it is full of surprises and uncertainty.
The problem is that inaccuracy in long-term projections can carry far greater costs, including the loss of public welfare. One of the most urgent areas is projecting the impact of climate change on the economy. The recent disaster in Sumatra shows starkly the cost of our inability to read nature's warning signs.
In this risk-laden landscape, mathematical modeling acts as a virtual laboratory for testing policy without paying for risk with real costs. It converts uncertainty into measurable risk. Through computer simulation, we can predict climate change and its economic impact, transforming policy from reactive to proactive, adaptive, and well-targeted. Integrated Assessment Models (IAMs), for example, bridge human economic activity and Earth systems, answering "what if" questions such as what happens if temperatures rise by 1.5°C. Those answers become the basis for figures in strategic documents such as Nationally Determined Contributions (NDC). Without them, climate commitments are merely empty promises on paper.
Indonesia has begun to adopt this path, integrating Low Carbon Development into the National Medium-Term Development Plan (RPJMN) 2020 to 2024. Bappenas used system dynamics and agent-based models to simulate environmental policy, with the main goal of cutting greenhouse gas emissions by 27.3% by 2024 while maintaining growth.
Even so, a model remains only a tool; institutional capacity must come first. The Sumatra floods of late November are bitter proof of how wide the gap between model and reality can be. Although an early warning (Cyclone Senyar) had been issued well in advance, weak mitigation and low awareness meant it was ignored. The result was heartbreaking: 1,189 lives lost and economic damage reaching Rp68 trillion to Rp200 trillion. This was not merely a natural disaster but a systemic failure, from relentless deforestation, to a cut of more than 50% in the National Disaster Management Agency's 2025 budget, to a warning that stalled at the bureaucratic level.
The failure should be a turning point for seeing how modeling integrated with real action can save lives and protect economic potential. Japan's G-Cans project cut flood duration drastically when Typhoon Hagibis struck in 2019, averting losses of up to US$1.76 billion. The Netherlands achieved similar results with its "Room for the River" strategy, and Singapore through smart dams and precise hydrological simulation. World Bank data reinforces that modeling is a strategic investment, not just a cost: improving early warning systems in developing countries could save 23,000 lives per year and generate up to US$30 billion in economic benefits.
The challenge of formulating sound economic policy is increasingly real. Uncertainty is rising, technology is changing rapidly, and climate impacts are growing more complex. Mathematical modeling should serve as a diagnostic tool for economists, much like a CT scan or MRI for a doctor. It lets policymakers see what is invisible, quantify hidden risks, and dissect the potential consequences of each decision, so they can formulate policies that minimize long-term damage.
Author: Muhamad Nagib Alatas
This article also appears on Katadata. Read the original here.
