Clean Water, Longer Life: Evaluating China’s Rural Drinking Water Safety Program
Drawing on two decades of village-level data, the study finds that China’s Rural Drinking Water Safety Program significantly increased rural longevity, with measurable gains in mean age at death emerging about five years after implementation and strengthening over time. The health benefits were largest in poorer villages with worse initial health, showing that large-scale safe water investments can both extend life expectancy and reduce rural health inequalities.
Conducted by researchers from the Key Laboratory of Geographical Process and Space Analysis Simulation of Hubei Province and the College of Urban and Environmental Sciences at Central China Normal University, together with scholars from the School of Economics and the School of Business Administration at Zhongnan University of Economics and Law and the Agricultural Economics Division of the Bangladesh Rice Research Institute, a study delivers a rare long-term assessment of how safe drinking water infrastructure reshapes population health in rural China. Using more than two decades of village-level data, the research evaluates the Rural Drinking Water Safety Program (RDWSP) as one of the world’s largest public investments linking water security to longevity.
From unsafe sources to a national water safety drive
The paper situates the RDWSP within China’s historical struggle to provide safe water to its rural population. For much of the twentieth century, rural households relied on untreated rivers, springs, and shallow wells, exposing residents to widespread waterborne disease and chronic health risks. Although incremental improvements began in the 1980s and 1990s, more than 70 percent of rural residents still lacked reliable access to safe drinking water at the start of that decade. This context prompted the Chinese government to launch the RDWSP between 2005 and 2015, investing roughly 282 billion yuan to expand centralized, treated, piped water systems. By the mid-2010s, hundreds of millions of rural residents were connected to water supplies meeting increasingly strict national quality standards, marking a decisive shift from crisis management toward long-term water safety.
Measuring health through the lens of longevity
Rather than focusing on short-term indicators such as diarrhea prevalence or child mortality, the study adopts the mean age at death at the village level as its core health outcome. This choice reflects an effort to capture cumulative health effects over time, acknowledging that longevity responds slowly to improvements in living conditions. Drawing on panel data from 360 villages across 31 provinces from 1995 to 2017, the authors can compare health trajectories before and after RDWSP implementation, while controlling for demographic, land-use, and socioeconomic factors. This long observational window allows the analysis to move beyond immediate correlations and toward a deeper understanding of how infrastructure investments translate into lasting health gains.
Short-term signals and long-term transformations
The econometric results reveal a consistent pattern. In the short term, villages with a higher proportion of households accessing safe drinking water show modest but statistically significant increases in mean age at death. While these immediate effects appear small, the authors emphasize that even incremental improvements in longevity reflect meaningful additions to health capital. The long-term effects are far more pronounced. Villages experiencing larger increases in safe drinking water coverage during the RDWSP period record significantly higher mean ages at death in subsequent years. On average, a one-percentage-point increase in access to safe drinking water between 2005 and 2015 is associated with an increase of about 0.03 years in mean age at death. Crucially, these benefits do not emerge immediately: event-study results show no significant gains in the first four years after implementation, with clear improvements appearing around the fifth year and strengthening thereafter. This delayed pattern underscores the cumulative nature of health improvements stemming from environmental infrastructure.
Unequal beginnings, unequal gains
The study also uncovers important distributional effects. Using quantile regression analysis, the authors find that the RDWSP delivers its largest health benefits in villages that initially had lower mean ages at death, where baseline health conditions were poorest. In these communities, improvements in water safety translate into disproportionately larger longevity gains. By contrast, villages already enjoying relatively high mean ages at death see smaller or statistically insignificant effects. This suggests that the RDWSP not only raised overall health outcomes but also helped narrow health inequalities within rural China, directing the greatest benefits toward the most disadvantaged areas.
Water infrastructure as long-term health capital
In its discussion, the paper frames these findings within health capital theory, arguing that safe drinking water functions as a productive investment rather than a simple welfare expense. Improved water quality reduces acute waterborne diseases in the short run and lowers cumulative exposure to chemical and microbial contaminants over time, thereby reducing chronic illnesses such as cardiovascular disease and certain cancers. The sustained gains in longevity documented in the study challenge concerns that large-scale rural infrastructure spending yields limited returns. Instead, the RDWSP emerges as a durable investment with lasting health and equity benefits. While acknowledging limitations, including reliance on village-level averages rather than individual mortality data, the authors conclude that China’s experience offers a powerful lesson for other developing countries: sustained investment in safe drinking water can generate long-term, inequality-reducing improvements in population health and contribute meaningfully to inclusive and sustainable development.
- FIRST PUBLISHED IN:
- Devdiscourse

