Comparing the trajectory of thermal and renewable power capacity in 2004 feels a bit like squinting through a dusty window at a landscape in transition. The numbers are there, mostly, but they don’t always clarify as much as one hopes. ISIC 3510—Electric power generation, transmission and distribution—holds both stories, the legacy of coal-fired giants and the uncertain beginnings of renewables, all bundled together under one industrial code. For an economist or policymaker, the challenge isn’t just to draw a line between them, but to make sense of why one grows as the other slows, and how policy intentions are (or are not) mirrored in the data.
The first move is to pull together the registry of all ISIC 3510-coded entities for the year in question. Most countries maintain lists—some neat, others a bit makeshift—of electric power producers and their generating assets. This is where the lack of granularity first shows up. The code covers everything: century-old coal stations, run-of-river hydro, solar arrays barely out of pilot stage. Sorting this out means extra work. Project-level documentation—sometimes found in energy regulator filings or company annual reports—helps assign individual plants to categories. For coal-fired stations, this is usually straightforward. Their output is often large, their location known, their operators steady over decades. Renewables, by contrast, are newer, their projects sometimes small, distributed, or folded into portfolios where the boundaries blur.
Once each firm’s assets are tagged as “thermal” (mostly coal, in this context) or “renewable” (wind, solar, biomass, or hydro, depending on national definitions), output figures become the next puzzle piece. Here, a patchwork of sources must be stitched together. Grid operator statistics, market reports, and government energy balances give the big picture. For more granular analysis, public filings—annual reports, investor presentations, or sometimes even environmental disclosures—supply unit-level data. Sometimes the numbers line up beautifully. Other times, a plant goes unreported, an operator consolidates figures, or an asset quietly changes hands and drops out of view.
The temptation to smooth out these inconsistencies is strong. Yet, the jagged edges of the data often say as much as the headline totals. For example, a sudden spike in renewable capacity from a handful of small firms may indicate a generous feed-in tariff policy, while a plateau in thermal output, despite continued investment, could signal regulatory or market resistance.
Aligning these output figures with national energy policies is its own balancing act. Policy documents rarely use ISIC codes. Instead, they speak of targets, incentives, and desired mixes—20% renewables by 2010, a cap on new coal after 2005, priority grid access for solar, and so forth. The trick is to place the data in dialogue with these intentions. A simple timeline, marking policy shifts against firm-level capacity changes, often reveals the lag between decree and reality. Some policies bite quickly; others take years to move the needle. Sometimes, a flurry of investment arrives in anticipation, only to fall off as the policy environment matures or loses momentum.
Cross-sectional comparisons add more texture. Looking at multiple countries—or even regions within a country—highlights how context matters. Two nations with identical ISIC 3510 registries can show very different expansion patterns if one relies heavily on coal subsidies while the other aggressively courts wind developers. The firm-level data, when aligned with announced policies, opens up room for these more nuanced readings.
There’s also the matter of reporting conventions. Some operators list installed capacity; others report net generation. Capacity factors vary widely—coal plants might run near full throttle, while solar or wind installations fluctuate seasonally and by region. Comparing raw output figures without adjusting for these factors risks distorting the picture. Yet, over-adjustment—stripping away every anomaly—can leave the analysis bloodless, too abstracted from the messiness of real-world transition.
Researchers should document every assumption and classification step, even the awkward or ambiguous ones. Tagging a plant as “renewable” might mean something very different in Denmark than in South Africa. A firm’s capacity figure could reflect planned, permitted, or actually commissioned assets, depending on the reporting source. Even the sequence in which new policy initiatives roll out has a bearing on the timing of capacity shifts.
ISIC 3510-coded data, handled with care and context, makes it possible to trace the shifting balance between thermal power and renewables in the early 2000s. The lines don’t always divide neatly, and the numbers occasionally resist telling a simple story, but when read against the backdrop of policy, market, and technology, they begin to reveal the undercurrents shaping energy’s future.