When crisis hits—like it did during the pandemic—governments are pressed to act fast, and act smart. The instinct is to throw a wide safety net over entire sectors, hoping no jobs slip through. But the reality is always more granular. Not every corner of manufacturing is hit the same way, and not every job is equally at risk. For policymakers and economists trying to get stimulus right, the ISIC codes offer a rare kind of clarity—one that lets you see, and support, the economy as it actually works.
ISIC codes break manufacturing down into fine-grained subcategories. Take ISIC 2710: Manufacture of electric motors, generators, transformers. Or ISIC 2100 for pharmaceuticals, ISIC 2821 for agricultural machinery, and so on. Instead of blanket support, governments can use these codes to map which sub-sectors are seeing the sharpest output declines, where layoffs are most acute, and where targeted relief might keep businesses afloat just long enough to survive the downturn.
The process begins by analyzing output and employment data at the ISIC subcategory level. National statistics offices, industry associations, and even payroll tax data can be coded by ISIC, showing which segments have suffered the biggest drops. Sometimes, what you find is counterintuitive—a subsector that looks healthy in aggregate may hide deep trouble in a particular niche, or vice versa. For example, demand for electric motors (ISIC 2710) may collapse if automotive production stalls, but surge if renewable energy projects are booming.
Economists can then use these insights to design proportional support measures. If output in ISIC 2821 falls by 30%, while ISIC 2100 is down only 5%, it makes sense to weight stimulus accordingly. Job retention schemes—like wage subsidies or payroll support—can be scaled to match the actual exposure, rather than using a flat percentage across manufacturing. This not only gets help to where it’s needed most, but it also avoids wasting resources on segments already recovering or less affected.
The work isn’t just technical; it’s strategic. By engaging with industry groups—who know the on-the-ground situation but may not always have the data discipline of ISIC coding—policymakers can cross-check findings, spot gaps, and adjust quickly as new information comes in. And as the situation evolves, repeat analysis means support can be ramped down or redirected, following the data rather than politics or pressure.
Of course, there are limits. Not every business fits neatly into a single ISIC code, and the codes themselves may lag behind new technologies or hybrid manufacturing models. Data reporting is sometimes slow, and by the time declines show up in official statistics, the worst may already be over. Still, the discipline of using ISIC subcategories forces a level of precision—and a humility—that blanket measures just can’t match.
Optimizing stimulus by ISIC isn’t about picking winners or losers. It’s about acknowledging that the economy is a mosaic, not a monolith. By targeting support to where it’s genuinely needed, policymakers can help ensure that when the next crisis comes, the tools for response are sharper, faster, and—most important—fairer to those most at risk.