Korea moved the agencies, the tax base followed, regional output did not

The spread is the finding: one host in ten on the headline output metric before controls, none in ten after them, seven in ten on local tax receipts ^[10]. And the spread has a plain mechanism. Payroll arrives, buildings and residents arrive with it, and public administration is itself counted inside regional output, so the public-sector line moves close to arithmetically ^[1]. Nothing in that chain requires a private firm to open anything.

The 40-kilometre comparison ring is what does the work. It exists to strip out improvement the whole surrounding region was getting anyway, so that what remains can be credited to the relocation rather than to the decade ^[4]. North Chungcheong had the conditions you would specify if you wanted the spillover story to hold: a relatively large site out of new residential land development, closeness to the Seoul metropolitan area, and highway and rail access ^[6]. Its neighbours moved too, and the apparent gain did not survive the comparison ^[7].

Worth noticing which lists the province appears in. North Chungcheong is in the group with a statistically significant real GRDP gain and, together with Busan and Ulsan, in the group without a statistically significant local tax gain ^[14]. The two channels are not tracking the same thing, which is consistent with the researchers' own reading that the province's advantage was land and location rather than tenants ^[6].

Lee's conclusion is conditional rather than negative: effects that start in the public sector spread only when private companies and investment come alongside the agencies and get connected to local industry ^[11], and the recommendation is a parallel package that redefines the agencies' role in regional industrial ecosystems, pulls in private investment, and widens the base for attracting firms and startups ^[12]. That conditional is the part the data cannot test. What was delivered was relocation, and relocation on its own is what got measured. As the researcher put it, in GRDP terms the effect was generally negligible, which points to limits on public agencies as a basis for self-sustaining regional growth ^[8].

Eighteen calendar years of indicators ^[13] is a long enough window that waiting is no longer a defence of the policy. The fiscal result is real and it is worth something: seven of ten host areas ended up with more revenue to spend ^[9]. It is simply a smaller purchase than the one the strategy is sold as, and it should be priced as the smaller purchase. The study is single-authored and, in English, so far reported by en.sedaily.com ^[2]; the method, not the volume of coverage, is the reason to take it seriously.