Current university policies regulate multi-sector careers through blunt quantitative caps, restricting external engagement without assessing its cross-sector value. This regulatory approach persists largely because the distinctive scientific value of individual "boundary-spanners" remains empirically undemonstrated at scale. Addressing this dual policy and empirical blind spot, we analyze 14.6 million publications to compare single-sector, team-based cross-sector, and boundary-spanner-led research. We provide evidence that boundary-spanner-led papers achieve significantly higher impact than single-sector papers (+8% in matched samples; OR = 1.20 for top-10% papers). Also, boundary-spanners achieve equivalent rates of top-tier research as cross-sector teams. Most importantly, we identify a phenomenon: sectoral breadth (diversity of sector types) yields substantially larger positive citation returns than institutional dispersion (total affiliations). Moreover, only institutional dispersion exhibits an inverted-U relationship, with marginal returns turning negative beyond 7-8 concurrent positions. These findings challenge the utility of blanket affiliation restrictions, calling for a transition from blind quantitative caps to qualitative frameworks prioritizing engagement diversity.
Tong et al. (2026) studied this question.