Sumitronomics 4.0: Digital Downstreaming – A Momentum Not to Be Missed
In his works on industrialisation, Sumitro Djojohadikusumo stressed the importance of leveraging comparative advantages, building future-oriented upstream industries, and creating structural linkages between upstream and downstream sectors (Djojohadikusumo, 1986). For Sumitro, without cross-sectoral linkages, economic independence is merely a statistical illusion.
Several decades later, the same principle urgently applies to the digital economy. Every second, hundreds of millions of Indonesians generate vast data flows—from online shopping clicks, travel histories, to content preferences. Yet, much like raw nickel ore once exported, this digital wealth mostly flows to servers in Silicon Valley, Shenzhen, or Singapore without domestic processing.
Yet, if Indonesia dared to halt raw nickel exports in 2020 to build domestic industry, why not do the same for data—the most valuable commodity of the 21st century?
Before discussing data downstreaming, empirical evidence from natural resource downstreaming must be highlighted. The raw nickel export ban initially sparked criticism, with the EU filing a WTO complaint and global economists labelling it protectionist.
Yet, five years later, results speak clearly. In 2024, downstream nickel products generated approximately $33.9 billion in exports—a dramatic surge compared to the raw export era. Indonesia now controls over half of global refined nickel production, with reserves of around 55 million tonnes out of the world’s total 130 million tonnes (USGS, 2024; Ministry of Energy and Mineral Resources, 2025).
This success aligns with Hartwick’s rule (1977), which states that profits from non-renewable resources must be reinvested into productive capital—here, through smelters, battery supply chains, and downstream industries.
This momentum confirms Sumitro’s upstream-downstream linkage principle is not theoretical dogma but an operational strategy delivering tangible benefits with measured political costs.
Data Colossus Still Selling Raw
Indonesia is a global data giant. A Google, Temasek, and Bain report (2025) projects Indonesia’s digital economy to reach $180-340 billion by 2030. Yet, much like nickel’s past, the highest value from this data flow is captured by foreign entities.
This phenomenon, termed ‘data colonialism’ by Couldry and Mejias (2019), continues colonial extraction logic through algorithms and global platforms.
Mordor Intelligence (2026) forecasts Indonesia’s cloud market to surge to $5.5 billion by 2031, growing at 14.32% annually. However, digital revenue is still dominated by foreign hyperscalers such as Amazon Web Services, Microsoft Azure, Google Cloud, and Alibaba Cloud. Around 66.05% of the domestic cloud market in 2025 consists of public deployments, largely served by foreign tech giants.
Mazzucato (2018) warns markets do not neutrally set value—governments must actively shape where value flows. Without strategic moves, the digital economy risks falling into extractive institutions (Acemoglu & Robinson, 2012), where digital wealth is continually extracted from the domestic ecosystem.
Reality shows the largest profits in the digital value chain are locked in algorithm and platform tiers, not raw data collection. If Indonesia settles for being a raw data extraction site, this structural imbalance will never shift.
Three Pillars of Digital Downstreaming
Digital downstreaming cannot be resolved by a single policy. Coordinated interventions across three data value chain pillars are needed, drawing lessons from nickel’s success.
The first pillar is processing infrastructure. Like smelters refining ore into ferronickel, Indonesia needs massive computing capacity to transform raw data into high-value AI models and services.
Critical initial steps have begun with the National Data Centre (PDN) in Cikarang, though full operation awaits security assessments. However, this effort is insufficient as most commercial data centres remain controlled and operated by foreign tech giants.
Genuine digital downstreaming demands genuine domestic ownership and operational control—not merely placing servers or physical infrastructure in the country. In this context, the World Bank (2025) also stresses the importance of digital compatibility standards to ensure interoperability and reduce foreign tech dependence.
The second pillar is algorithmic capability. Nickel becomes battery cathodes not just due to smelters but skilled electrochemists. Data downstreaming requires talent capable of building sector-specific machine learning models. The target of 9 million digital talents by 2030 needs re-evaluation on quality.
Currently, most training programs produce platform operators—not algorithm architects. World Bank research found only around 10% of jobs in East Asia and the Pacific involve AI-complementary tasks, far below advanced economies that