China's successful eradication of absolute poverty by 2020 has shifted the focus to relative poverty, particularly in rural areas. Traditional income-based poverty measures are insufficient to capture the multidimensional nature of poverty, as highlighted by Sen's work. The Alkire and Foster (2011) multidimensional poverty index (MPI) provides a more comprehensive framework. The widespread adoption of smartphones in China, exceeding 90% of rural households, presents a potential tool for poverty alleviation. However, the impact of smartphone usage on poverty reduction remains debated. Some argue that smartphones increase income through improved market access and information, while others express concerns about exacerbating the digital divide. This paper aims to clarify this complex relationship by expanding the MPI to include income and labor dimensions, constructing a more nuanced measure of poverty for post-poverty households. It also investigates the mediating role of social capital, considering its various dimensions. The study focuses on rural households in Jiangxi province, a key area in China's poverty alleviation efforts.

Existing literature presents contrasting views on the impact of smartphone usage on poverty reduction. Some studies highlight the positive effects of increased income through improved access to markets and information. For instance, Zhuo et al. (2023) found that smartphones increase farmers' incomes by overcoming spatial barriers. Other research suggests a 25-30% increase in supplementary wage income through smartphone usage (Van and Van, 2009; Panteli et al., 2019). Furthermore, smartphones facilitate job searching, non-farm employment (Zeng et al., 2023), and entrepreneurial activities (Mack et al., 2017). Studies also demonstrate the role of smartphones in disseminating agricultural information and enhancing the well-being of migrant workers (Aker et al., 2016). Liu et al. (2021) found a positive relationship between digital information technology, social capital, and multidimensional poverty reduction. Conversely, other studies raise concerns about the digital divide, suggesting that smartphone use may exacerbate inequalities (Tayo et al., 2016). Acılar (2011) noted the difficulty in bridging the digital gap between rural and urban areas. Deichmann et al. (2016) point to limited digital literacy and IT access as barriers to benefiting from the digital dividend. This study seeks to build upon these existing studies by offering a more comprehensive analysis of the relationship between smartphone usage, social capital, and multidimensional poverty.

This study uses data from a survey of 382 rural households in five counties and districts of Jiangxi province, China, conducted in 2020. The sample included households that had successfully escaped absolute poverty. A multidimensional poverty index (MPI) was constructed using ten indicators across five dimensions: health (medical expenses, health insurance), education (years of education), income (per capita disposable income), living standards (electricity, cooking fuel, floor quality, assets, per capita housing area), and labor force (proportion of labor force in household). The A-F approach was used to calculate the MPI, categorizing households into Vulnerable Multidimensional Poverty Index (VMPI), General Multidimensional Poverty Index (GMPI), and Extreme Multidimensional Poverty Index (EMPI) groups. Smartphone usage was the primary independent variable, with social capital (measured by social networks, social trust, social participation, and social support) as the mediating variable. Control variables included individual characteristics (gender, age, education, marital status) and household characteristics (household size, number of laborers, household income). Multivariate ordered logistic regression models were used to analyze the data, with robustness checks using computer usage as a proxy for smartphone usage and bootstrap methods to assess mediation effects.

The study's key findings are as follows: 1. **High Prevalence of Multidimensional Poverty:** The overall incidence of multidimensional poverty was 42.8%. However, this was significantly higher (51.8%) among households without smartphones compared to households with smartphones (24.5%). 2. **Education, Labor, and Health as Primary Drivers:** Education (37.8%), labor force (29.7%), and health (20.4%) were the main contributors to multidimensional poverty. 3. **Smartphone Usage Reduces Poverty:** Smartphone usage significantly reduced VMPI (57.6%), GMPI (52.6%), and EMPI (5%). The effect was less pronounced as poverty severity increased (EMPI). 4. **Social Capital's Mediating Role:** Social capital fully mediated EMPI reduction through smartphone usage (91.67%) and partially mediated VMPI (14.09%) and GMPI (20.84%) reduction. This suggests that smartphones' positive impact on poverty is channeled through the enhancement of social capital. 5. **Inverted U-Shaped Relationship:** The study found an inverted U-shaped relationship between the stringency of poverty criteria and the multidimensional poverty index (MPI). As the poverty thresholds increased, the MPI initially rose before subsequently declining. 6. **Robustness:** Robustness tests using computer usage as a proxy confirmed the main findings, reinforcing the impact of internet access on poverty reduction.

The findings demonstrate a clear link between smartphone usage and multidimensional poverty reduction in rural China. The significant reduction in all three MPI categories (VMPI, GMPI, EMPI) supports the idea that smartphones empower rural households. The diminishing effect as poverty severity increases suggests that while smartphones are beneficial, they are not a panacea for the most extreme forms of poverty. The crucial mediating role of social capital highlights the importance of community networks and social support in translating technological access into tangible improvements in well-being. The study's findings contribute to the literature by providing empirical evidence of the multifaceted impact of smartphones, emphasizing the significance of considering social capital in poverty alleviation strategies. The inverted U-shaped relationship between poverty criteria stringency and MPI suggests a need for targeted interventions focused on the most vulnerable populations.

This study provides strong evidence that smartphone usage contributes significantly to reducing multidimensional poverty in rural China, particularly through its influence on social capital. The findings emphasize the importance of incorporating both technological access and social capital enhancement into poverty reduction strategies. Future research could explore the specific mechanisms through which smartphones build social capital and examine the long-term sustainability of these effects. Further studies could focus on specific vulnerable groups, such as adolescents and women, and consider the psychological aspects of poverty.

The study's limitations include its reliance on cross-sectional data, which limits causal inference. The sample, while representative of post-poverty households in Jiangxi province, may not be generalizable to other regions or contexts. Furthermore, the study's measurement of social capital, while comprehensive, may not capture all its nuanced aspects. Future longitudinal studies with larger, more diverse samples would be beneficial to further investigate the complex interplay between smartphone usage, social capital, and multidimensional poverty reduction.