Facilitating fertility decline through economic development: a principal-agent analysis of local bureaucratic incentives in China's fertility transition

The paper addresses why political science has had limited explanatory power in fertility transition compared with economics and sociology, despite observable fertility fluctuations linked to political actions. It posits that prevailing work overemphasizes direct, coercive fertility policies (e.g., the one-child policy) and assumes mechanical implementation, overlooking indirect bureaucratic processes and incentives. The authors propose that, under a principal-agent framework within China’s centralized yet delegated governance system, local officials pursue career advancement by achieving centrally mandated outcomes. They argue that officials can fine-tune fertility outcomes through strategies that avoid social instability, especially by promoting economic development that indirectly lowers TFR. The study aims to clarify the bureaucratic mechanisms linking promotion incentives, economic development, and fertility decline, thereby integrating political demography with economic and sociological explanations of the fertility transition.

The review integrates economic, sociological, and political-demographic perspectives. Economic and sociological theories attribute long-run fertility decline to rising opportunity costs of childbearing, human capital investment, urbanization, education, and shifting norms (Becker; Caldwell; Lesthaeghe). Yet historical fluctuations in TFR often align with political interventions rather than smooth economic trends. Political demography has been underdeveloped by focusing narrowly on direct policies and policy decisions rather than implementation mechanisms. Principal-agent theory (PAT) explains central-local relations in China: the center (principal) sets outcome-based targets; local governments (agents) possess information advantages and face promotion incentives tied to quantifiable performance, risking incentive distortions (Holmstrom & Milgrom; Li & Zhou). Promotion tournaments prioritize economic performance but may also include fertility targets. Prior work documents strategies local officials use to boost growth and critiques of the promotion model (resource constraints, factionalism, reassignment). Synthesizing this literature, the authors hypothesize: H1a/H1b linking promotion incentives to economic development and promotions; H2a/H2b linking incentives and fertility decline to promotions; H3 that radical birth-control campaigns reduce promotion prospects due to instability; and H4 that officials facilitate fertility decline through economic development.

Data: A novel provincial panel (province-year) covering 1980–2000 links provincial Party secretaries and governors (resumes from the Lingnan Local Officials Database, 1978–2008) to socioeconomic indicators (CNKI statistical database) and TFR measures (1980–1990 from Chen & Coale 1993; 1991–2000 from National Bureau of Statistics of China and East-West Center 2007). Xinjiang and Tibet were excluded due to unique policies; officials with tenure under one year were dropped. Final sample: 141 provincial party secretaries and 156 governors, processed as person-year panels (n = 1153; models report 1086 observations after exclusions and lags). Provincial fixed effects control time-invariant differences in family planning regulations (notably second-child approval rules). Measurement: Dependent variables include logged TFR (age-specific TFR; robustness with PPR-based TFR) and ordinal political turnover (0 termination/demotion/resignation, 1 sideway move, 2 promotion), following Li & Zhou (2005). Explanatory variables for promotion motivation are critical age and tenure: proximity to age 59 (with age 60 typical exit), age squared; proximity to turnover year; dummies for 1 and 2 years before turnover. Performance measures: within-tenure per capita GDP growth and industrialization growth (industrial added value/GDP), and within-tenure TFR decline; squared TFR decline proxies radical birth-control campaigns and tests for nonlinearity (inverted U). Mediators: per capita GDP and industrialization rate. Covariates: For TFR models—urbanization, rural female labor share, divorce and marriage rates; For promotion models—official characteristics (age 59+ indicator, education) and economic performance controls (growth in social investment, electricity, agricultural machinery power, CPI). Research strategy and models: 1) System two-step GMM dynamic panel models predict ln(outcomes) from promotion motivation, lagged outcomes, mediators/covariates, and province fixed effects (Arellano-Bover; Roodman). 2) Ordered logistic regression (OLM) models predict promotion odds from economic and fertility-decline performance and controls. 3) Partial mediation analysis: promotion motivation → economic outcomes; then motivation + economic outcomes → fertility outcomes, assessing whether economic development mediates the effect on TFR. Robustness: Replace age-specific TFR with PPRS TFR for key specifications; results consistent. All models include province fixed effects to absorb time-invariant policy heterogeneity in birth regulations.

- Promotion incentives and economic performance: Proximity to age 59 and to turnover year is positively associated with higher per capita GDP and higher industrialization rates (e.g., each year closer to age 59 associates with approximately +10 RMB per capita GDP and +0.1 percentage points in industrialization; proximity to turnover year associates with ~+20 RMB per capita GDP). Economic performance raises promotion odds: 1% higher per capita GDP growth associates with about 1.2% higher odds of promotion (OLM). - Promotion incentives and fertility decline: Age effects are U-shaped for TFR with a turning point at 59; being closer to age 59 and being in the year before turnover associate with lower TFR. Within-tenure TFR decline increases promotion odds: each 1% decrease in TFR associates with about 2.6% higher promotion odds; relative to predecessor, each 1% additional decline associates with ~2.4% higher promotion odds. - Nonlinear effect and radical campaigns: The relationship between TFR decline and promotion odds is inverted-U with a turning point around a 37% TFR decline within a term; larger declines reduce promotion odds, consistent with penalties for destabilizing radical campaigns. - Mediation via economic development: Economic development negatively associates with TFR. A 10,000 RMB increase in per capita GDP associates with a 5% reduction in TFR; a 1 percentage point increase in industrialization rate associates with a 0.152% reduction in TFR. Promotion-incentivized officials thus facilitate fertility decline indirectly by boosting economic development. - Overall: Results support H1a, H1b, H2a, H2b, H3, and H4. Officials facing promotion pressure improve economic metrics and reduce TFR; both performances aid promotion up to a stability-constrained threshold, and economic development serves as an effective, lower-risk pathway to fertility decline than radical campaigns.

The findings demonstrate that bureaucratic incentives embedded in China’s outcome-based target system align local officials’ behavior with central goals on both economic growth and fertility decline. By modeling the central-local relationship as a principal-agent problem, the study clarifies how quantified promotion incentives drive proactive behaviors at critical ages and tenure points. Rather than relying on direct and potentially destabilizing birth-control campaigns, officials rationally adopt economic development as a low-risk strategy that indirectly reduces fertility through higher opportunity costs, longer education, and labor-market dynamics. This reframes political demography’s role: economics and sociology explain long-run downward trends in fertility, while political processes account for fluctuations and strategic fine-tuning. The inverted-U relationship between TFR decline and promotion odds underscores central constraints prioritizing social stability: excessive, rapid declines signal risky campaigns and are penalized. Integrating political demography with economic-driven theories, the study shows how bureaucratic processes mediate fertility transition, enhancing explanatory power for observed patterns in China’s 1980–2000 experience.

Using provincial panel data (1980–2000) and a principal-agent framework, the study shows that promotion-incentivized local officials pursued both economic development and fertility decline, with economic growth serving as an effective mediator for reducing TFR. Key contributions include: identifying critical age/tenure effects driving performance pushes; establishing that economic and fertility performance both improve promotion odds but that excessive fertility decline (over ~37% per term) is penalized; and demonstrating that rising GDP per capita and industrialization are associated with lower TFR. The work advances political demography by detailing implementation mechanisms—how bureaucratic incentives enable fertility transitions—thus bridging political and economic-sociological paradigms. Future research could incorporate richer covariates (e.g., ethnic minority shares, migration), leverage improved administrative and survey data to address measurement concerns, extend analyses to later periods and subprovincial levels, and explore heterogeneity by policy regimes and local market conditions.

- Omitted covariates: Lack of controls for ethnic minority proportions and interprovincial migration due to data inaccessibility may leave residual confounding, though authors argue small shares in 1980–2000 likely limit bias. - Data quality: Potential underreporting/misreporting in the 1990s fertility data; authors use the best-available provincial estimates (NBS and East-West Center) and note scholarly debate on the extent of underreporting. - Generalizability: Focus on China’s 1980–2000 provincial leadership and institutional context; results may not generalize to different bureaucratic systems or later periods without institutional similarity. - Timing and mechanisms: Economic development effects on fertility may operate with lags that vary across locales; the study’s tenure-focused design may not capture all dynamic pathways. - Measurement proxies: Radical campaigns inferred via squared TFR decline; while informative, it remains an indirect proxy for coercive enforcement intensity.