South Korea, once praised as an "economic miracle" for enjoying high growth of 8 to 9 percent annually through the 1980s, now stands before the fear of a "compressed fall." Since the mid-1990s, the growth rate has fallen by 1 percentage point every five years, and it is now trapped in a range around 1 percent. At this pace, the country will face the brutal reality of "zero" growth within a few years.
The cause of the declining growth rate is falling productivity. In economics, the key to raising productivity has been the "creative destruction" described by Joseph Schumpeter, and its engine is the startup. Peter Howitt, a professor at Brown University in the United States who visited Korea in May, won last year's Nobel Prize in Economics for his work linking Schumpeterian creative-destruction growth theory with endogenous growth theory. His core argument is that only countries equipped with the institutional and financial infrastructure that allows innovative startups to be founded easily and rapidly forces out existing firms with declining productivity can enjoy long-term productivity growth. As capital and labor move toward high-productivity sectors, the economy grows. The positive correlation between growth rates and startup rates supports this. In particular, to break the "K-shaped polarization" structure in which gaps between large and small firms and income inequality are deepening, revitalizing startups is not a choice but a necessity.
Indicators showing the dynamism of our economy are flashing warning signs everywhere. Korea's startup rate not only falls short of that of the United States or China but has also been on a continuous downward slope since the 2000s. The number of companies founded annually, which reached 1.48 million in 2020, fell to 1.14 million in 2025. Technology-based startups, which should serve as the backbone of the economy, also declined from 24,000 to 22,000 over the same period.
The government's recently announced "Startups for All Project," aimed at reversing this trend, is welcome in that it encourages everyone to take on innovation and spreads the results to the regions. But the compass of policy must shift away from the quantitative obsession with "how many were added" toward the qualitative question of "how productive the startup is." Academia's recent focus on "young firms" rather than startups themselves is in the same vein. Korea's productivity currently stands at 80 percent of the average of the Organization for Economic Cooperation and Development (OECD), and the service sector does not even reach 70 percent. The particularly low productivity in the service sector is because the proportion of small business owners in retail, food, and lodging is excessively high. If the recent rise in startups is concentrated in subsistence businesses rather than manufacturing or knowledge services, it could only fuel overcrowding among small business owners rather than productivity growth through innovation.
For the government's "Startups for All" policy to bear the fruit of genuine growth, three supplementary measures must be put in place. First, the center of gravity must shift from subsistence small-business startups toward "innovative tech startups" that directly lead to productivity gains. Second, growth support measures must be carefully designed so that new firms can rapidly "scale up" after founding. In a structure where firms fail to cross the death valley and close early, neither a healthy startup ecosystem nor productivity growth can be expected. Third, startups must serve as a breakwater against regional extinction. Regions without jobs cannot avoid population outflow and extinction. To make the birth of companies start in the regions themselves, a "university-company-local government" cooperation model must be put into operation immediately.
At the crossroads of a compressed fall, what we need is not quantity but quality. We look forward to a refined evolution of policy that encourages true "innovative startups" capable of restoring dynamism and lifting the potential growth rate once again.
