Internet Use and Subjective Well-Being in China

The iGDP indicator, which measures a country’s Internet economy, employs the expenditure approach to calculate GDP. Generally, it sums up all activities associated with the creation and use of Internet networks and services (Woetzel et al. 2014).

Valkenburg and Peter (2007) identify some potential mechanisms of Internet communication and well-being via social networks among Dutch adolescents aged 10–17.

Each item on the 20-item Young Internet Addiction Test (YIAT) is measured on a 5-point scale ranging from “1 = not at all” to “5 = always”.

Pathological use of the Internet is evaluated by the self-rated 20-item Young Internet Addiction Test, a 5-point scale with scores ranging from 20 to 100, which are then used to group addiction severity into three categories: normal = 20–49, moderate = 50–79, and severe = 80–100 (Lam and Peng 2010). In Lam and Peng’s (2010) study, only 10 students scored 80 points or higher, so the researchers use a binary variable (1 = severe and moderate and 0 = normal) in their analysis.

The 2010 CFPS encompasses 14,960 Chinese households and 42,590 individuals (0 < age ≤ 110), excluding Hong Kong, Macao, Taiwan, Xinjiang, Qinghai, Inner Mongolia, Ningxia, and Hainan. Detailed information about sampling design is available in Xie (2012).

We mainly focus on the Chinese Internet users until 60 years old because Internet use in older ages (>60) becomes very limited and heterogeneous with regards to motives. Also note that the rate of Internet use participation is only 1.5 % among people aged above 60 years olds in our sample.

Specifically, the 25 provinces/municipalities/autonomous regions in the CFPS are Beijing, Tianjin, Hebei, Shanxi, Liaoning, Jilin, Heilongjiang, Shanghai, Jiangsu, Zhejiang, Anhui, Fujian, Jiangxi, Shandong, Henan, Hubei, Hunan, Guangdong, Guangxi, Chongqing, Sichuan, Guizhou, Yunnan, Shaanxi and Gansu (Xie 2012).

Because a considerable share of our respondents do not use the Internet and the decision to do so may be related to reported SWB, our analysis may be subject to sample selection bias.

These unreported results are available from the authors upon request.

Nevertheless, it is important to emphasize that causality may be the opposite: individuals that are less satisfied with life may perceive the Internet as an important means for diversion and distraction.

Although we do not test for causality, our results could imply that heavy Internet users may be less happy and more depressed because they feel guilty for neglecting important responsibilities such as work, housework, or taking care of family, especially given that, in spite of rapid modernization, the Chinese still believe in strong traditional values. Interestingly, Davis et al. (1992) find that, in the workplace, the US people’ intentions to adopt computer technology are primarily influenced their perceptions for the usefulness of the computer technology (utilitarian values) for improving job performance and productivity, and less by the degree of enjoyment (hedonic values), which is further echoed by Igbaria et al. (1995) among Finnish professional and managers. Similarly, using data from an electric Webpage survey in Singapore, Teo et al. (1999) show that Internet users use the Internet primarily because they perceive the Internet to be more useful to their job tasks, and only secondarily because it is enjoyable and easy to use, suggesting that perceived usefulness is more important that perceived ease of use and perceived enjoyment in affecting Internet usage.

Because the 2010 CFPS provides information on the substitution between Internet use and time spent watching TV (by asking whether the latter has increased or decreased substantially since Internet use began), we also run an ordered probit model with TV watching as the dependent variable. We find that time spent on Internet use decreases the probability of spending more time watching TV, implying that longer Internet exposure may reduce viewing time. The estimation results, although not reported here, are available from the authors upon request.

As highlighted by Pénard et al. (2013), relative to Internet use participation, the intensity of Internet use (e.g. frequencies of Internet use) is a better proxy for individual’s exposure to the Internet. More importantly, it is able to capture the “intent-to-use” in a better fashion.

Bessiere et al. (2010), using a 2000–2002 national US panel survey of 740 individuals, find that health-related Internet use is associated with increased depression. Likewise, Forton et al. (2007) also confirm that higher frequencies of Internet use for meeting people, socially experiment, and participating in chatting rooms are related to more depressive symptoms among US undergraduate students.

Selfhout et al. (2009) provide the evidence that for Dutch adolescents with low-quality friendships, spending more time on Internet activities is associated with increased depression and social anxiety. Similarly, Valkenburg and Peter (2007) find that Internet communications are correlated with decrease in well-being among Dutch adolescents.