The Daily Lives and Job Searches of SNEP

Appendix

The Determinants of IT and SNS Usage

An ordered probit analysis was performed to identify the factors that determine the frequency of communicating via e-mail and retrieve and acquire information from the Internet.

When filling out the survey in 2011, respondents indicated how many days during the preceding year they used e-mail and the Internet: 1–4 days, 5–9 days, 10–19 days (1 day a month), 20–39 days (2–3 days a month), 40–99 days (1 day a week), 100–199 days (2–3 days a week), or 200 days or more (at least 4 days a week). With use of the ordered probit estimation method, the factors are verified behind the responses in the more frequently answered categories.

Then a dummy variable for SNEP is added as an independent variable, as well as dummy variables for gender, age, educational background, annual household income, the presence of a person requiring long-term care in the household, the opportunity for receiving medical treatment or care, and the size of city by population. The estimated results are presented in Table 3.4.

Table 3.4 Determinants of use of e-mail and information retrieval and acquisition (2011)

According to the table, the coefficient for the solitary non-employment dummy is significantly negative for both e-mail and information retrieval . Those non-employed persons who use e-mail and retrieve information online tend to be young, come from a well-educated background, and live in a large city. Meanwhile, the UMNEP who live in households earning less than 4 million yen per year had the lowest e-mail and Internet use rates.

In Table 3.5, a solitary non-employment dummy is divided into two groups: individual-type SNEP and family-type SNEP. The table clearly shows that family-type SNEP demonstrate significantly low rates of e-mail use and information retrieval. When separating the SNEP into those who participated in sports, travel, or volunteer work during the preceding year and those who had not, SNEP who had not participated in such activities were also less likely to have used e-mail or searched for information on the Internet.

Table 3.5 Determinants of use of e-mail and information retrieval and acquisition (2011)

Table 3.6 shows the results of the examination of the influence solitary non-employment had on the frequency of PC and smartphone use in the 2016 STULA. It is easy to confirm that solitary non-employment had significantly negative impacts on PC and smartphone use. Divided into individual-type and family-type SNEP dummies, both of the coefficients were significantly negative, and the magnitude was slightly larger for individual-type SNEP than for family-type SNEP.

Table 3.6 Effects of solitary non-employment on use of PC and smartphone (2016)

The Determinants of Gaming Behaviors

As hikikomori, and the disconnected society in general, continue to garner more and more public attention, the issues of “Internet addiction ” and “Internet dependence ”—conditions in which people shut themselves up in their own homes or rooms to indulge in online gaming for long periods of time—are also drawing concern.

In the “Leisure Activities” section of the STULA, the list of hobbies and amusements includes an item for “Playing TV games or PC games (include use of portable game machines).” Respondents had to indicate whether they had played games and, if so, how often they had done so over the course of the preceding year.

Based on the data for this survey item, an ordered probit analysis was conducted using gaming frequency as the explained variable and solitary non-employment, gender, age, academic background, household income, receiving medical treatment or care, the presence of a person who requires long-term care in the household, and city class as independent dummy variables. The surveys from 2006 and 2011 were focused on, because they correspond to the period when online gaming became popular. Table 3.7 presents the findings on how SNEP fit into the gaming picture. The table also includes estimates of gaming-related figures for the SNEP population, broken into different segments: individual-type SNEP vs. family-type SNEP and SNEP who participate in sports, travel, or volunteering experience during the preceding year vs. SNEP who did not.

Table 3.7 Effects on annual use of gaming behaviors

It is apparent from the estimated results that the effect of being SNEP was insignificant in 2011 and actually had a significantly adverse impact on game use in 2006. These results could reflect the fact that SNEP demonstrated very low levels of Internet use in 2006, thereby limiting their opportunities to participate in online gaming.

Dividing the SNEP population into individual and family types also revealed some interesting findings. In both 2006 and 2011, individual-type SNEP exhibited significantly low levels of gaming frequency. Family-type SNEP, on the other hand, were significantly averse to gaming in 2006 but displayed insignificant results in 2011. SNEP who did not engage in any sports, travel, or volunteer work over the preceding year demonstrated significantly low levels of gaming frequency for both survey years.

Viewed from a larger perspective, these results run contrary to the notion that SNEP are Internet addicts who are preoccupied with gaming. At present there is little evidence to support the argument that the socially isolated, solitary non-employed population is growing due to a surge in the number of non-employed persons who are indulging in gaming by themselves.

The Determinants of Time Use

The “Time Use” section of the STULA requires respondents to report on their lives over a reference period of two consecutive days, by selecting activities from among 20 categories. By selecting from these 20 pre-coded leisure activity categories, respondents indicate what activities they participated in, and for how long, over each 48-hour period in 15-minute blocks.

Tables 3.8 takes the four categories with the highest distribution of overall time from Table 3.1 in order to estimate the corresponding determinants via the ordinary least squares method. The explained variables were the 2-day averages for each time use category from the 2016 STULA. For independent variables, dummies for solitary non-employment, gender, age, educational background, household income, the presence of a person who requires long-term care in the household, receiving medical treatment or care, and the size of city by population were used. The weather conditions during the two reference dates and the corresponding days of the week are also added to the list of independent variables.

Table 3.8 Determinants of spending time for leisure activities (2016)

Solitary non-employment, as the estimates show, contributes to a significantly longer amount of time spent watching TV , listening to the radio, reading newspapers, reading magazines, practicing hobbies , pursuing amusement, resting , relaxing , and sleeping , even when variations among the different attributes are controlled. No matter how thoroughly the other attributes are accounted for, TV, etc., is one of the activities with the starkest differences between solitary and non-solitary non-employed persons: on average, SNEP watch 42.6 more minutes of TV per day than non-solitary non-employed persons do. Sleeping, hobbies, and amusements follow the same pattern, with SNEP spending at least 15 min more than their non-solitary counterparts on these activities.

For Table 3.9, the same estimation is used as for Table 3.8, replacing the solitary non-employment dummy variable with dummy variables for individual-type SNEP and family-type SNEP. As shown in Table 3.1, individual-type SNEP—and family-type SNEP as well—spent significantly longer amounts of time on TV, etc., sleeping, hobbies, rest, and amusements, compared with non-solitary non-employed persons.

Table 3.9 Determinants of spending time for leisure activity: individual or family-type (2016)

The Determinants of Job-Seeking Activities and the Motivation to Work

In order to more rigorously examine the effects of solitary non-employment on individuals’ job-seeking activities and their desire to work, a probit analysis was conducted, using a variety of independent variables. The investigation focused on three types of dependent variables.

The first dependent variable was whether the person in question wanted to work and was, in some capacity, actively looking for a job. The second variable was whether individuals wanted to find a job, regardless of whether they were engaged in any job-seeking activities. The third variable was whether the individuals had engaged in any learning or training for getting a job over the preceding year. Considering that “NEET ” is an abbreviation for “Not in Education, Employment, or Training,” the investigations encompassed not only the subjects’ attempts to find work but also any education or training that prepared them for potential employment (Social Exclusion Unit 1999). By looking at the effects of a person’s job-seeking activities, desire to work, and job-oriented learning, the relationship between NEET and SNEP can be analyzed as directly as possible. Independent variables are the same as those in Table 3.4.

Table 3.10 shows the estimates obtained from the 2006, 2011, and 2016 data, respectively. In each year, being SNEP had a statistically significant negative effect on that person’s job-seeking activities , desire to work , and job-oriented learning or training . Although job-seeking activities and desire to work may have varied according to gender, age, educational background, and a wide range of other attributes, solitary non-employment had a detrimental effect on achieving future employment, even when accounting for the differences in these observable attributes.

Table 3.10 Determinants of seeking a job, desire to work, and learning or training for getting a job

Table 3.11 Determinants of seeking a job, desire to work, and learning or training for getting a job: individual or family-type

In 2006, 2011, and 2016, receiving medical treatment or care significantly restricted job-seeking activity; thus, the growth of the nonlabor force might be due not only to Japan’s aging society but also to the health problems of the non-employed population. In addition, living with a person who requires long-term care limited job-seeking activity in 2006 and 2016.

For Table 3.11, some of the estimations were further performed, replacing the solitary non-employment variable in Table 3.10 with dummies for individual-type SNEP and family-type SNEP. These results again underscored that the family-type SNEP have the lowest levels of job-related activity, job-related awareness, and job-oriented learning or training. Controlling for other various attributes, it is apparent that the differences between individual-type SNEP and non-solitary non-employed persons in job-seeking activity, desire to work, and job-oriented learning were insignificant in most cases.

Finally, it is also possible to see from Table 3.11 that the absolute value of the marginal effect of being a family-type SNEP on job-seeking declined from 2006 to 2016. Comparing the marginal effects of family-type SNEP between 2006 and 2016, it seemed to have the least impact on the desire to work and job learning or training in 2016. Combined with the trend shown in Fig. 3.5, this shows that Japan’s non-employed persons tend to be discouraged from job seeking and desiring work, regardless of their solitary situations. These tendencies toward lost interest in work should be more carefully examined, using the data from STULA and other current and future labor statistics.