The researchers measured Talent Policy Attention, or TPA, from the volume of talent-related news coverage in each city and year. Public response was captured through a talent policy sentiment-interaction index, TPSI, built from messages posted on the Message Board for Leaders, a government-run platform where citizens can address officials directly. The index combines the sentiment composition of those messages with the intensity of institutionalized interaction, so it reflects not only how favorable public expression is but also how actively residents engage with authorities on the policy.

Using two-way fixed-effects models with a quadratic term, the analysis estimates a turning point at roughly 94 talent-related news reports per city-year in this sample. Below that level of attention, higher coverage is associated with a lower TPSI; above it, the association reverses and becomes positive. The authors stress that the design is observational at the city level, so the estimates describe associations rather than causal effects.

A closer look at the components of the index locates the reversal in participation rather than in sentiment polarity. Measures built on sentiment composition alone do not turn within the observed range of attention. In other words, sustained news attention is associated with more talent-policy messages being posted, not with a more favorable balance of expression. Cities with heavy coverage do not appear to shift public opinion toward praise; they appear to draw more residents into institutionalized channels of communication. The pattern proved robust across alternative specifications of the TPSI index and also held for the participation dimension taken on its own, suggesting the nonlinear relationship is not an artifact of one particular way of scoring public response.

The study also probes the conditions under which the association takes different forms. Mediation and heterogeneity results are consistent with possible conditioning roles for digital economy development and public service capacity. Cities with more developed digital economies and stronger public service capacity may be better positioned to convert sustained policy attention into constructive online engagement, although the observational design means these mediating pathways remain suggestive rather than demonstrated.

The findings carry practical weight for city governments that treat talent attraction as a communications campaign. Simply increasing the volume of talent-policy news may, at moderate levels of attention, coincide with a less favorable public response, while only at higher levels of sustained coverage does the association turn positive. The results imply that visibility and engagement follow a more complicated dynamic than a simple more-is-better rule. The study’s scope also defines its limits: the panel covers Chinese cities over a decade in which talent competition intensified nationally, and the Message Board for Leaders captures a specific, institutionalized form of online expression rather than the full landscape of social media reaction. The estimates describe associations at the city level and cannot establish that news attention causes changes in public response.

Subject of Research: Talent policy attention and online public response: a nonlinear association across 232 Chinese cities

Article Title: Talent policy attention and online public response: a nonlinear association across 232 Chinese cities

Article References: Li, K., & Wang, G. (2026). Talent policy attention and online public response: a nonlinear association across 232 Chinese cities. Humanities and Social Sciences Communications. https://doi.org/10.1057/s41599-026-09310-5

Image Credits: AI Generated

DOI: 10.1057/s41599-026-09310-5

Keywords: Talent, policy, attention, online, public, response, nonlinear, association, across, Chinese, cities, scientific research