Bridging the gap: the effect of rural e-commerce development on internal income inequality in China

ABSTRACT

The burgeoning sphere of rural e-commerce has garnered considerable acclaim for its digital economy dividends, yet the literature presents a gap in evaluating its repercussions on household internal income inequality. Regarding the Rural E-commerce Demonstration County (REDC) program implemented in batches as a quasi-natural experiment, we employ the staggered difference-in-differences (SDID) method to explore causal relationships between rural e-commerce development and internal income inequality. Results show that rural e-commerce development significantly reduces internal income inequality in China. Using the Bacon decomposition method and exclusion of potential spillover effects, we confirm the reliability of our SDID estimates. We find that this positive effect is explained by non-farm employment and entrepreneurial activities, information advantages, and infrastructure development. Furthermore, we reveal the inclusive characteristics of rural e-commerce development but are influenced by the potential digital divide. Our study enriches the research field on the benefits of e-commerce from a new perspective of alleviating income inequality and offers pragmatic insights for harnessing e-commerce tools as vehicles for attenuating income disparity and fostering social equity in developing countries.

KEYWORDS:

• E-commerce development
• internal income inequality
• relative deprivation index
• inclusive characteristics
• rural China

JEL CLASSIFICATION:

• R11
• O38
• I31

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability

The authors do not have permission to share data.

Notes

¹ Data source: http://dzsws.mofcom.gov.cn/article/ztxx/ndbg/201805/20180502750562.shtml.

² It is also known as the Kakwani index.

³ Data source: http://www.mofcom.gov.cn/article/zt_dzswjnc/lanmufive/201705/20170502572784.shtml.

⁴ The total income of rural households in the NFRS database includes household business income, wage income from local employment, wage income from outside employment, rental income, etc., but does not cover remittances.

⁵ Although we focus on rural households, there are numerous migration populations due to China’s large population base. Therefore, our sample includes some householders who live in rural areas but have a non-rural hukou.

⁶ Since the observation years are 2009–2017 and the policy pilot years are 2014–2017, the range of values for k is [−8,+3]. Based on the fact that the first pilots were in 2014 and the latest observation year was in 2017, the maximum value of k can be calculated to be + 3. Similarly, the last pilots were in 2017 while the earliest observation year was in 2009. Therefore, the minimum value of k is −8.

⁷ Both periods −7 and −8 are pooled to period −6 to avoid full covariance because we strictly control the household FE and the reduction in observations in the treatment group when k was less than −6. In Figure 3, we use −6+ for this purpose.

⁸ Work type is a binary variable assigned the value of 1 if the rural householder is employed in a non-farm job, and 0 otherwise. Given the lack of direct statistics on household entrepreneurial activities within the NFRS database, we introduce a second binary variable, Income sources. This variable is set to 1 if the principal source of a rural household’s income stems from private business operations. To reflect the scope of entrepreneurship in rural areas, we adopt a macro-level variable Company numbers, which signifies the number of enterprises per 1,000 individuals in the village at the year-end.

⁹ We meticulously gathered panel data from 2,097 counties from 2010 to 2020. Our objective is to determine whether the REDC program can promote non-farm employment and entrepreneurial activities at the macro level. The findings, elaborately presented in Appendix C, provide substantial evidence supporting our thesis. They confirm that the implementation of REDC program is indeed instrumental in motivating rural residents to pursue non-agricultural jobs and embark on entrepreneurial endeavours.

¹⁰ The former is a continuous variable, while the latter two are binary variables. They are set to 1 if a household owns a TV or a mobile phone, and 0 otherwise.

¹¹ The first two variables are binary; they are assigned to 1 if the rural household has access to running water or possesses air conditioning or heating, and 0 otherwise. Electricity usage represents the number of households within a village that are connected to the electrical grid.

Additional information

Funding

The work was supported by the National Natural Science Foundation of China [72373043,No. 72304102]; National Social Science Fundation of China [23&ZD112]; Philosophy and Social Science Program Discipline Co-construction Project of Guangdong Province [No. GD23XYJ68].