In China, demographic projections have informed gradual increases in the statutory retirement age. Photo: IC PHOTO

Predictive research is an important tool for understanding social dynamics and formulating policies. Given the complexity of socioeconomic phenomena, discrepancies between predictions and actual outcomes are inevitable. Some critics therefore argue that predictive research is “useless” and ineffective in informing practice. This is, however, a misunderstanding. Predictive research does not pursue absolute accuracy; rather, it aims to reveal trends and risks, offer frameworks for navigating uncertainty, facilitate informed decision-making, as well as advance academic work and policy practice. 

Value of predictive research 

In the face of skepticism, it is important to reevaluate the significance of predictive research. First of all, a scientific understanding of the nature of predictive research is essential. A common misconception is that predictive research should accurately forecast every detail of the future. In reality, it presents potential future trends by simulating diverse scenarios, analyzing historical data, and integrating expert opinions. The essence of scientific prediction lies not in delivering exact outcomes, but in recognizing patterns and potential trajectories within intricate systems, using limited data and informed assumptions. Predictive outcomes are influenced by numerous factors, including complex variables, unforeseen events, flawed model parameters, and data quality. 

For example, the “World Population Prospects” report published by the United Nations projects population trends over the coming decades by analyzing data such as fertility and mortality rates. While its methods are well-developed and widely adopted, its predictions can still deviate from actual trends at times due to unforeseen events such as pandemics. 

The core value of predictive research lies in trend identification and risk management. Despite the criticism that the inherent errors in predictive models limit their utility, revision of models and assumptions is, in fact, common in scientific research. Studies on climate change, for instance, have long predicted rising global temperatures. Although early forecasts lacked accuracy, ongoing research and continuous improvements in climate models have brought predictions increasingly in line with observed realities. These advances have provided the scientific basis for global emissions reduction policies and environmental protection efforts. 

The value of predictive research also resides in its ability to foresee socio-economic transformations and offer forward-looking recommendations. Studies show that many countries, including China, are experiencing a trend toward smaller households dominated by nuclear family structures. This shift directly affects consumption patterns as well as demand for energy, transportation, housing, and community services, posing challenges to resource allocation and social security systems. To address these challenges, it is necessary to prioritize the design and construction of small and medium-sized residences while also promoting the development of emerging industries such as smart homes, food delivery services, short-term rental platforms, and telemedicine. 

Furthermore, predictive research plays a vital role in guiding policy formulation and optimization. In Europe, early demographic forecasting allowed many countries to anticipate the challenges of an aging population, leading to adjustments in retirement policies, labor markets, and pension systems. Similarly, in China, demographic projections have informed gradual increases in the statutory retirement age and improvements to healthcare and elderly care services to address the growing pressures of an aging society. 

In addition, predictive research is instrumental in evaluating policy outcomes. Since the impacts of policies often take time to materialize, predictive models can analyze medium- and long-term trends to identify potential shortcomings, enabling policymakers to make timely adjustments. For instance, research indicates that lifting birth restrictions alone may not immediately reverse declining birth rates. This insight has prompted the Chinese government to optimize its birth support policies, including improving maternity leave benefits and creating more female-friendly workplaces, thereby enhancing policy effectiveness from multiple angles. 

Evaluation of predictive research 

In recent years, predictive research has made remarkable progress in uncovering complex social issues and informing policy recommendations, yet there is still room for improvement. The introduction of methods such as sensitivity analysis, multi-scenario forecasting, and multidimensional policy simulation can offer policymakers more diverse and comprehensive perspectives. The application of new technologies such as big data analytics and artificial intelligence can considerably enhance forecasting efficiency and accuracy. Furthermore, developing interdisciplinary predictive models and continuously updating their parameters ensures that these models remain responsive to real-time socioeconomic changes, thereby enhancing their adaptability and foresight. 

The value of predictive research should be assessed along several dimensions. First, its ability to identify long-term trends and potential changes is essential in determining its significance as a guide for the future. Second, to ensure the credibility of predictions, models and analytical methods must meet the basic standards of scientific research, including rigorous data collection, well-founded assumptions, and transparency of results. Third, the practical relevance and feasibility of predictive findings are crucial to policy-making. Even when forecasts deviate from actual outcomes, they retain value as long as they can effectively inform decision-making. 

In conclusion, predictive research should be evaluated objectively, with a focus on its contributions to addressing real-world issues rather than simply on its limitations. As methodologies improve, interdisciplinary collaboration deepens, and new technologies are deployed, predictive research has the potential to provide stronger scientific support for sustained social development. 

Gu Danan is a researcher at the United Nations Population Division. 

Edited by WANG YOURAN