Investment Attraction and the Prevention of Rural Child Left-Behind Status in Chinese Counties
Year: 2026
Type: article
Abstract: Using a balanced panel of 892 counties in 10 labor-exporting Chinese provinces over 2015–2022, this paper estimates how county-level investment attraction affects the number of rural left-behind children within county borders. Two-way fixed effects estimates indicate that a 1 percentage point rise in realized in-vestment funds relative to GDP lowers the county left-behind children (LBC) rate among compulso-ry-education students by 0.181 percentage points. A Bartik instrument built from base-year industry composition and national sectoral investment growth yields a two-stage least squares estimate of −0.381 (Kleibergen–Paap F = 221.81); a difference-in-differences design around the approval of development zones in 158 counties yields −1.651 percentage points, with no detectable pre-treatment trends in the event study. Estimators built for staggered adoption reproduce this magnitude (Callaway–Sant’Anna −1.592; Sun–Abraham −1.614), and the approval coefficient measures the effect of the zone approval event, not the effect of one percentage point of investment intensity. Sub-sample estimates locate the effect in western counties, high out-migration counties, and counties whose investment inflows lean toward labor-intensive industries, with coefficient gaps of 2.3–3.2 times across groups. The effect runs through local non-farm employment (Sobel z = −6.65, accounting for 31.8% of the total effect) and the out-migration share of the rural labor force (Sobel z = −3.04, 11.8%); the fiscal-livelihood channel receives only marginal support at the 10% level, and the three channels jointly explain 46.1% of the total effect. The share of children with both parents away falls by 0.134 percentage points per percentage point of investment intensity, which places the change in the form of left-behind status associated with the largest cognitive losses. The evidence supports writing employment-oriented investment attraction into county-level policy portfolios for LBC prevention and calibrating investment quality, meaning labor intensity, hiring per unit of capital, and capacity to employ women, rather than investment volume alone.
Institutions: Lincoln University College, SEGi University
Language: English
Field: Social Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W7212157296