Skip to main content

Digital inclusion and wellbeing in New Zealand

Executive Summary

We examine two main questions relating to internet (and other ICT) access:

  • Which groups have a lower likelihood of being digitally included in New Zealand (and why)?
  • How does digital inclusion relate to waiora/wellbeing?

In examining the first question, we pay particular attention to the situation for Māori and Pasifika relative to other ethnic groups.

Existing research examines aspects of who is digitally excluded. It is, however, important also to examine how access relates to people’s wellbeing alongside their access and use. Our second question starts to address whether internet access is beneficial for specific communities.

Methodology

We use four large-scale surveys of New Zealanders that include information on internet availability. Some of the surveys also include information on availability of other ICT related items and on internet use.

The surveys are:

We place most emphasis on NZCVS and NZES as they are the most recent of the surveys. We also consider PISA as it includes adolescents as well as containing added information on how adolescents use the internet. The surveys are each well sampled but all the figures must nonetheless be treated as having some degree of sampling error.

Access to the internet

A number of groups are prone to relatively low access to the internet, including:

  • people living in social housing
  • people with disabilities
  • Pasifika
  • Māori
  • people living in larger country towns (10,000 to 25,000 people)
  • older members of society, particularly those aged over 75 years
  • unemployed people and those not actively seeking work.

The first two of these groups – those in social housing and people with disabilities – appear to be particularly disadvantaged with respect to internet access. Pasifika students (in 2015) also reported substantially lower rates of internet access than did students of other ethnicities.

Just 69% of those living in Housing NZ (or local equivalent) social housing report having access to the internet, compared with 91% reporting access across all respondents (in the 2017 NZES).

Only 71% of people with disabilities report having access to the internet (in the 2017 NZES). In the 2018 NZCVS, 17% of people with disabilities indicate having no internet access compared to the full sample where just 5% have no internet access.

These large gaps in internet access for those who live in social housing and for people with disabilities are potentially amenable to policy interventions. Most social housing is owned by the state, local authorities or NGOs. The social housing provider could take the initiative to install WiFi (or other technologies) to enable internet access by tenants. Provision of such infrastructure may be considered of similar importance to provision of water, sewerage and electricity to these tenants. Such provision is also likely to improve internet access rates for Pasifika students.

Similarly, many people with disabilities are already subject to some form of care by state agencies or NGOs. These authorities may consider enabling internet use for their clients as a key intervention to improve the opportunities for those with disabilities to connect with the rest of society.

People with disabilities are also at greater risk than others from an internet violation (such as a virus infection or other internet interference). Other at-risk groups include individuals who are not actively seeking work, unemployed, Māori, Pasifika, younger people, and those who are studying.

Wellbeing and internet use

We investigate the association between various wellbeing indicators and internet use. As we have used cross-sectional data we cannot draw causal conclusions on the nature of these associations.

Our key findings are:

  • NZCVS (adult) data indicate that those who do not have internet access tend to have lower subjective wellbeing than those who do have access.
  • NZES (adult) data show a similar relationship (using a proxy variable for wellbeing) when we do not control for other (e.g. demographic) factors, but we find no relationship once we control for these other factors.
  • NZES data show that people without internet access are less engaged in civic activities such as voting in general elections and making submissions to government.
  • PISA (adolescent) data indicate that those without internet access tend to have lower subjective wellbeing than those with access (which may reflect family circumstances).
  • PISA data also indicate that as internet use on weekdays outside of school increases, students’ subjective wellbeing declines; once daily internet use exceeds about two hours, we find no positive association between internet use and wellbeing.

The PISA data show that 15% of 15-year olds (including 27% of Māori students) report using the internet for more than 6 hours per day on a weekday outside of school, while over half report more than two hours’ use.

Recommendations and policy considerations

We recommend that policy consideration be given to two particularly at-risk groups: social housing residents, and individuals with disabilities. A range of policy interventions already addresses issues faced by each of these groups. There appears to be a strong case that interventions be extended to enabling internet access for these individuals.

Those who work with youth (and their family members) may wish to give consideration to assessing the effects of prolonged use of the internet by adolescents. Our associative results – while not establishing a causal link – highlight a potential concern relating to wellbeing outcomes for those with prolonged internet use. We recommend further investigation of the wellbeing effects of extended use of the internet – both for adolescents and, if the data is available, for children and adults.

We also recommend further analysis of emerging and future PISA, NZCVS and PIAAC data relating to internet (and ICT) access and use. These analyses will be able to leverage the links that these surveys will shortly have to Statistics NZ’s Integrated Data Infrastructure (IDI). By linking the survey results to prior characteristics of the surveyed individuals and of their localities (via the IDI), researchers will be able to better control for personal and other traits that affect both wellbeing and internet (and other ICT) access and use.

1. Introduction

Advances in information and communications technologies (ICTs), including the internet, have led to advances in life expectancy, GDP, life satisfaction, and environmental sustainability (Bughin et. al. 2019). Access to the internet is fundamental to the effective digitalisation of New Zealand. Accordingly, the New Zealand government has set strategic priorities relating to the digital domain including internet access (Ministry of Business, Innovation & Employment & Stats NZ, 2019).

While there has been prior research on internet access and use in New Zealand (Strickland & Evans 2018, InternetNZ 2017, MBIE 2015, Statistics New Zealand 2018a, Statistics New Zealand 2017, Statistics New Zealand 2018b and Digital Inclusion Research Group 2017), there has been little research examining the relationship between internet access and wellbeing either in New Zealand or globally. According to Bughin et. al. (2019) ICTs (including access to the internet) may be neither good nor bad per se, but ICT access unequally impacts different parts of the population. We provide information both on internet access and use across population segments in New Zealand, and on the relationship between internet use and wellbeing.

Our focus is on two main questions relating to ICT access, with our principal focus being on internet access:

  • Which groups have a lower likelihood of being digitally included in New Zealand (and why)?
  • How does digital inclusion relate to waiora/wellbeing?

In examining the first of these questions, we pay particular attention to the situation for Māori and Pasifika relative to other ethnic groups.

We draw on a variety of domestic data sources to explore these questions. The most recently surveyed datasets include the New Zealand Crime and Victims Survey (NZCVS, surveyed in 2018) and the New Zealand Electoral Survey (NZES, surveyed in 2017). We also analyse the most recently available dataset relating to adolescents, the Programme for International Student Assessment (PISA, surveyed in 2015). Results of the 2018 PISA survey are scheduled for release in late 2019.

In the Appendix we also present results from the Programme for the International Assessment of Adult Competencies (PIAAC, surveyed in 2014/15). Each survey includes questions about internet access; the NZCVS, PISA and NZES have questions that we use also to examine the relationship between internet access and subjective wellbeing (waiora).

Using this data, we provide information about the digital divide in New Zealand. Our focus is on the difference between those who have internet access and those who do not. We find that Māori, Pasifika, those living in social housing, unemployed people, those not actively seeking work, disabled individuals, those living in larger country towns and older members of society are less likely to have internet access. The largest gaps in internet access are for those who identify as living in social housing, being disabled, unemployed, and/or in older age groups. These findings are broadly consistent across surveys (where similar information is collected).

Using PISA, NZCVS and NZES data, we infer the correlation between subjective wellbeing and internet access (and other forms of digital inclusion). The PISA data indicate that, while internet access is associated with higher subjective wellbeing among students, this effect reduces as more time is spent on the internet. The NZCVS data show that amongst adults, lower life satisfaction is associated with a lower likelihood of internet access. Unlike the PISA data, the NZCVS associations do not control for other characteristics of individuals. For the adult population in NZES, we again find a difference in (a proxy measure of) wellbeing between those with and without internet access when we do not control for other factors, but this relationship disappears once we control for demographic and other influences. However the NZES data show that those with internet access have higher rates of voting in general elections and are much more active in other forms of civic participation than are those without internet access. Thus internet access appears to be positively correlated with these aspects of social capital.

One feature of the NZCVS is that it provides data on which segments of the population are most at risk of having their computer infected by a virus or being otherwise interfered with; we refer to these as occurrences of ‘internet violation’. Māori and Pasifika are more at risk of internet violation than other ethnicities. In addition, people who are disabled, not actively seeking work and/or with low subjective wellbeing having heightened risk of internet violation.

The remainder of this report contains a brief literature review that highlights gaps in existing knowledge, a data section which provides a brief description of the datasets used and relevant questions from those datasets (with further information provided in the Appendix) and a methodology section.

We present our results relating to digital inclusion (principally related to internet access) in section 5 (with further details provided in the Appendix), and our wellbeing-related results are presented in section 6.

A concluding section discusses potential avenues for further research relevant to the promotion of digital inclusion policies in New Zealand.

2. Literature Survey

The literature on internet access in New Zealand deals with topics that include who has access to the internet, the quality of internet connection, how people use their internet and internet security. In this report, we are primarily interested in who has access to the internet and the relationship between internet and wellbeing.

There has been considerable research which looks at the digital divide, both globally and in New Zealand. One category of research looks at the groups in society that make most use of the internet. For example, French, Quinn & Yates (2018) find that education is a predictor of internet use in the UK. Studies have also found a relationship between age and use of the internet; a commonly found result is that internet use diminishes with age, especially for those aged 65 and older (Andrade et. al. 2017, Stephanie 2018, Smith et. al. 2016 and Auckland University of Technology 2018). A related issue, that it is difficult to answer with the data currently at hand, is whether this is an age effect or a cohort effect; for instance, it is conceivable that internet access for people aged over 75 may be commensurate with the rest of the population in one to two decades’ time as the current 50-60 year old age-group moves into later life.

Studies show that New Zealand has a relatively high proportion of people with access to the internet (Strickland & Evans 2018, InternetNZ 2017, MBIE 2015, Statistics New Zealand 2018a, Statistics New Zealand 2017, Statistics New Zealand 2018b and Digital Inclusion Research Group 2017). These studies indicate that further research is nevertheless still required to look at the relationship between internet connection and certain demographics, for instance the rural vs urban divide and internet access according to disability status. We provide a more in depth demographic breakdown here, using the demographic groups available in PISA, NZES, NZCVS and PIAAC.

One issue of policy interest for which data has been somewhat lacking in recent studies is analysis of internet access for Māori, Pasifika and other ethnic groups in New Zealand. A study by MBIE (2015) found that internet access for Māori was lower than for other groups in New Zealand, mirroring a prior finding by Greenbrook-Held & Morrison (2011) who also found low access rates for Pasifika.

Bughin et. al. (2019) examined the relationship between technology and wellbeing globally, highlighting the many benefits of access to the internet and other aspects of ICTs. One more cautionary study relating to adolescents in England is that by Przybylski and Weinstein (2017). They show that while a small amount of screen use (such as use of the internet) is associated with improved mental health, screen use of longer than one to two hours is associated with poorer adolescent mental health outcomes. There is little or no research exploring similar relationships in New Zealand. Indeed, perhaps the biggest gap in the internet-related literature for New Zealand is analysis of the relationship between internet access and wellbeing.

This report aims to close the gaps in the literature with regard to how internet access varies across demographic groups. We also provide evidence of the relationship between wellbeing, internet access and intensity of internet access in New Zealand. Related to wellbeing issues, we present new findings on the groups within society that are most at risk of internet crime, an area that has hitherto been understudied both in New Zealand and globally.

3. Data

The surveys primarily used for this analysis include the Programme for International Student Assessment (PISA, 2015), the New Zealand Electoral Survey (NZES, 2017) and the New Zealand Crime and Victim Survey (NZCVS, 2018). Each survey includes questions about internet access and each has questions that enable us to examine correlations of internet access with subjective wellbeing. In the Appendix, we also discuss the (more dated) Programme for the International Assessment of Adult Competencies (PIAAC, 2014) survey.

In this section, we briefly outline the three main surveys used. The Appendix provides further details on the survey questions used from each of these surveys.

PISA

PISA is an international survey of 15 year olds conducted by the Organisation for Economic Co-operation and Development (OECD), involving both OECD member and OECD non-member countries. It collects demographic and academic information on these students. The information includes questions on internet access and access to devices with internet access; in some countries it also includes questions on subjective wellbeing. We use the 2015 PISA survey, which is the most recent PISA survey with information that is publicly available. To make sure the results used are adjusted appropriately for population proportions, sample weights are applied to the descriptive statistics, supplied by PISA in their dataset. The demographic variables of interest include gender, parent post school education level, and ethnicity. There is a total of 4,520 observations in the 2015 New Zealand PISA dataset available. Surveys were administered at school with students entering their responses via computer.

As New Zealand did not include the question about subjective wellbeing (SWB) for students in its study, we use the 2015 PISA data for Great Britain to formulate a proxy SWB variable for New Zealand (see the Methodology section for the approach adopted).

NZES

NZES is a study which posts questionnaires across the country to randomly selected registered people with the right to vote in New Zealand. For each election since 1990, NZES has been conducted and within each survey there is a subsample of individuals who have answered the questionnaire in prior election years. We focus on the 2017 survey for the majority of descriptive statistics and for the wellbeing association but also use the longitudinal nature of the data, analysing internet access for individuals in 2011, 2014 and 2017. The demographic information we use from NZES includes ethnicity, gender, age, education, working situation, whether the person is self-employed, the type of area a person is residing in, voting activity, income and housing situation. The questions we consider for the longitudinal analysis remain the same across the 2011, 2014 and 2017 datasets. The sample size for the 2017 survey respondents is 3,455. The longitudinal sample size (across the 2011, 2014 and 2017 NZES respondents) is 536. For the descriptive statistics, we use the sampling weights provided in the datasets. We construct an indicator relating to (lack of) wellbeing based on a range of responses to questions in the survey (see the Methodology section for the approach adopted).

NZCVS

NZCVS is an annual survey that collects information about New Zealanders’ experience of crime. The dataset we use is from 2018; it surveyed 8,000 people aged 15 years and over (each by an interviewer). We focus on the survey’s question about whether an individual’s computer or internet enabled device had been infected or interfered with, to determine: (a) who has access to the internet, and (b) the groups of people who are most at risk of internet violation. We also relate these results to the reported life satisfaction across different demographic groups.

4. Methodology

Methodology applied to the data:

PISA

For each dataset used in this study, we created summary statistics of internet access or internet use broken down by demographic variables available in each survey. PISA has two internet variables, time spent on the internet and access to the internet. Using the ICT familiarity questionnaire we determined who had spent time on the internet at school on weekdays, and outside of school on weekdays and on weekends. For each category, we identified students who spent no time on the internet, and then grouped others into their respective reported online times (or into the ‘unknown’ internet group).

To determine internet access at home and at school, we again used the ICT familiarity questionnaire. An individual was defined as having internet access at home if they had internet connection at home or if they had a cell phone with internet access or both. An individual was defined as having access to the internet at school if they had internet-connected school computers or internet connection via a wireless network or both. For both home and school, these categories were then broken down to internet access is available and the student uses it, internet access is available and the student does not use it, and the student does not have internet access.

The demographic variables we used from PISA were gender and ethnicity. For gender, respondents had the choice of being male or female. We report the proportion of each gender according to time spent on the internet and internet access. For ethnicity, the data has Māori, Pacific, Asian, Other and Pākehā options. As students can say they are more than one of these options, we split the ethnicity categories into: Māori, Pacific, Asian, Other, Pākehā, Pākehā and Māori, all other combinations, and unknown. As with the gender statistics, the number and proportion for each ethnicity was reported for time spent on the internet and internet access. All results were weighted using the 2015 PISA sampling weights.

To analyse the correlation between wellbeing and internet access, we ran an ordinary least squares (OLS) regression of estimated subjective wellbeing (SWB) on internet use and other correlates using the 2015 PISA data from New Zealand. PISA contains a subjective wellbeing question but this question was not asked in New Zealand in 2015 (or in the 2018 survey). Instead, we used the United Kingdom (UK) wellbeing results to construct a proxy for New Zealand students’ SWB.

To do so, we first ran an OLS regression of UK wellbeing against a variety of variables within PISA that are hypothesised to be related to SWB, selecting variables which were statistically significant (in the UK wellbeing regression) at the 1 percent level. The variables selected were: home possessions, sense of belonging, value of cooperating, parental support, instrumental motivation and test anxiety. The UK regression is reported in the Appendix.

We created a proxy subjective wellbeing variable for New Zealand students using the UK regression coefficients applied to the corresponding New Zealand data. We then ran an OLS regression of this SWB proxy against ICT-related variables and student demographic variables.

ICT variables comprised: time spent on the internet at school on weekdays and outside of school on weekdays and on weekends, desktop at home, laptop at home, tablet at home, internet connection at home, cell phone with internet access at home, desktop at school, laptop at school, tablet at school, computer at school with internet, wireless internet connection at school, age first used a digital device, age first used a computer, and age first used the internet. Demographic variables comprised: age, gender, ethnicity, mother education attainment and father education attainment.

NZES

The NZES data was used to produce summary statistics relating to internet use together with regression results relating to the association between internet access and wellbeing. The summary statistics produced are for 2017 and we also utilise longitudinal data for 2011, 2014 and 2017. The longitudinal results are presented in the Appendix.

For the 2017 summary statistics, we examined internet access in relation to available demographic variables. Respondents were included in the internet access group if they had one or more of the following: internet at work, internet at home, internet on mobile or internet somewhere else and did not say they had no access to the internet. Respondents who only said they had no access to the internet were put in the no internet access group. All others were placed in the unknown internet group.

The 2017 summary statistics include the number and the proportion of people within each demographic who did or did not have access to the internet. Demographic groups of variables included ethnicity, gender, age, education, working situation, the type of area a person is residing in, and housing situation. Separately, we present summary statistics relating to internet access according to voting and other forms of civic participation. All results are weighted using the sample weights provided.

The longitudinal summary statistics detail the proportion of people according to age, ethnicity and work status who did or did not have access to the internet. In addition we show, for the longitudinal sample, the pattern of internet access for respondents over the past three waves of the survey (i.e. 2011, 2014, 2017). As expected, this reveals considerable persistence in internet uptake but also shows cases in which an individual loses internet access across time.

To analyse whether there was a correlation between wellbeing and internet access, we ran a regression between a dissatisfaction variable (i.e. a proxy for lack of wellbeing) and internet access using 2017 NZES data. The dissatisfaction variable was formed as the first principal component of 12 variables from the 2017 NZES that we believed could signal general dissatisfaction of an individual. The questions we used are shown in the Appendix.

Each of the scales from the questions were coded to be in the same direction for which 1 is least dissatisfied and 4 or 5 is most dissatisfied. All of the variables used in the analysis had positive loadings (as expected); the first principal component had an eigenvalue of 2.65, so explained 22% of the variance across the 12 variables. This suggests that while the derived variable is likely to be correlated with respondents’ generalised dissatisfaction it may not be a strong summary measure. Using OLS, we regressed the dissatisfaction variable against internet access plus a range of demographic variables: gender, age, self-employment, working situation, ethnicity, income, and education level.

NZCVS

The NZCVS 2018 data includes information on no internet access based on an internet related question. We defined no internet as those who answered not applicable to a question about whether their computer or internet connect device had been infected or interfered with. Not applicable came with the added explanation in the survey that no one in the house has owned a computer or internet-enabled device in the last 12 months.

The summary statistics show the proportion of people who answered this question with regards to life satisfaction and demographics. The demographic information used includes ethnicity, sex, age, household income, employment status and disability status. The results are not weighted and the counts have been randomly rounded to base three using the rules defined by Statistics NZ. (The results do not include respondents who responded don’t know or refused to answer.)

5. Results: Internet Access

PISA

We use the PISA data to examine internet access of 15-year olds in terms of time spent online at school and at home, plus internet access at school and at home. We break the statistics down by gender and ethnicity using sampling weights. Recall that the data refer to 2015, so some patterns may have changed in the interim.

Table 1 and Table 2 show access to the internet broken down by gender. We find a reported gender difference in access at school with males (76.54%) reporting less access than females (81.45%). There is little difference in access at home. Similarly, there is little difference between males and females in terms of time spent on internet outside of school. The notable difference is for time spent on the internet at school, with 80% of females, compared with 74% of males, spending time on the internet (see Table 3, Table 4, and Table 5).

Total numbers in all PISA and NZES tables are after applying sampling weights.

Table 1: Internet access at home by gender

  Yes and use it Yes but don't use it No No response Total
Female 85.99% 1.04% 0.89% 12.08% 26,966
Male 84.60% 1.04% 1.22% 13.14% 27,309
Total 85.29% 1.04% 1.06% 12.62% 54,274

Table 2: Internet access at school by gender

  Yes and use it Yes but don't use it No No response Total
Female 81.45% 4.49% 0.86% 13.20% 26,966
Male 76.54% 6.59% 1.95% 14.92% 27,309
Total 78.98% 5.55% 1.41% 14.07% 54,274
Table 3: Time spent on internet at school by gender
  No time Time spent No response Total

Female

5.61%

80.02%

14.37%

26,966

Male

10.08%

73.91%

16.01%

27,309

Total

7.86%

76.94%

15.20%

54,274

Table 4: Time spent on internet outside school on weekdays by gender
  No time Time spent No response Total

Female

1.92%

83.77%

14.31%

26,966

Male

1.77%

82.35%

15.87%

27,309

Total

1.85%

83.06%

15.10%

54,274

Table 5: Time spent on internet on weekends by gender
  No time Time spent No response Total

Female

1.94%

83.19%

14.87%

26,966

Male

1.65%

81.98%

16.37%

27,309

Total

1.79%

82.58%

15.62%

54,274

Table 6 and Table 7 report internet access at school and outside of school, broken down by ethnicity. Table 6 shows internet access at home. Pasifika students report far lower internet access rates at home (74%) than do students of all other ethnicities; Pākehā students have the highest rate of home access, approximately 6 percentage points higher than the Māori access rate. Those of “unknown” ethnicities appear to have an extraordinarily high “no response” to the internet questions. These are likely to be students who answer the survey questions incompletely and so are not discussed here.

Similarly, Pasifika students report much lower internet access at school than do other students (although the high “no response” rate for Pasifika students may influence the Pasifika results). Again Māori students lag Pākehā students in this respect, lagging by nine percentage points in school internet use. The lagging rates of particularly Pasifika internet access (and, to a lesser but still material extent, Māori access) at school – if reported accurately – is of special concern given these students’ comparative lack of internet access at home.

Table 6: Internet access at home by ethnicity
  Yes and use it Yes but don't use it No No response Total

Māori

85.05%

1.65%

3.98%

9.31%

4,215

Pasifika

71.62%

2.36%

1.58%

24.45%

3,545

Asian

85.25%

0.41%

0.44%

13.90%

6,492

Pākehā

90.98%

0.86%

0.75%

7.42%

28,901

Māori and Pākehā

88.98%

1.20%

0.93%

8.89%

4,391

Other

79.95%

0.00%

0.00%

20.05%

680

All other combinations

80.18%

1.80%

1.38%

16.65%

4,675

Unknown

10.02%

0.00%

0.00%

89.98%

1,373

Total

85.29%

1.04%

1.06%

12.62%

54,274

Table 7: Internet access at school by ethnicity
  Yes and use it Yes but don't use it No No response Total

Māori

75.42%

6.72%

3.69%

14.17%

4,215

Pasifika

66.40%

5.87%

1.24%

26.49%

3,545

Asian

78.63%

4.47%

1.75%

15.16%

6,492

Pākehā

84.60%

5.76%

1.25%

8.39%

28,901

Māori and Pākehā

82.35%

5.72%

1.15%

10.78%

4,391

Other

73.09%

4.73%

0.00%

22.18%

680

All other combinations

75.73%

5.68%

0.85%

17.74%

4,675

Unknown

8.98%

1.04%

0.00%

89.98%

1,373

Total

78.98%

5.55%

1.41%

14.06%

54,274

When we examine the amount of time spent on the internet (Table 8, Table 9, and Table 10), we observe the same patterns. Pasifika students are much less likely to report spending any time on the internet both at school and at home (on weekdays and weekends) than do students of other ethnicities. Māori students lag Pākehā students in spending any time on the internet whether at school or at home, with the gaps between use rates being about seven percentage points in each case.

Table 8: Time spent on the internet at school by ethnicity
  Time spent No time spent No response Total

Māori

75.12%

9.83%

15.05%

4,215

Pasifika

58.71%

10.16%

31.13%

3,545

Asian

76.55%

8.42%

15.03%

6,492

Pākehā

82.93%

7.78%

9.29%

28,901

Māori and Pākehā

80.11%

7.97%

11.92%

4,391

Other

69.42%

4.06%

26.52%

680

All other combinations

74.20%

6.35%

19.45%

4,675

Unknown

8.31%

1.71%

89.98%

1,373

Total

76.94%

7.86%

15.20%

54,274

Table 9: Time spent on the internet outside school on weekdays by ethnicity
  Time spent No time spent No response Total

Māori

82.62%

1.66%

15.72%

4,215

Pasifika

66.08%

3.34%

30.58%

3,545

Asian

83.62%

1.09%

15.29%

6,492

Pākehā

89.33%

1.45%

9.22%

28,901

Māori and Pākehā

85.20%

3.20%

11.60%

4,391

Other

75.70%

1.32%

22.98%

680

All other combinations

77.10%

3.70%

19.20%

4,675

Unknown

10.65%

0.00%

89.35%

1,373

Total

83.06%

1.85%

15.10%

54,274

Table 10: Time spent on the internet on the weekend by ethnicity
  Time spent No time spent No response Total

Māori

82.18%

1.70%

16.12%

4,215

Pasifika

64.76%

3.97%

31.26%

3,545

Asian

82.82%

1.08%

16.10%

6,492

Pākehā

88.98%

1.41%

9.62%

28,901

Māori and Pākehā

85.58%

2.72%

11.70%

4,391

Other

74.19%

1.32%

24.49%

680

All other combinations

76.71%

3.36%

19.93%

4,675

Unknown

8.61%

0.00%

91.39%

1,373

Total

82.58%

1.80%

15.62%

54,274

In the Appendix, we report additional tables using the PISA data relating to the amount of time spent on the internet per day by students (according to gender and ethnicity). As with the tables in this section, we include separate tables for internet use at school, internet use outside of school on weekdays, and internet use per day on weekends. Of those who responded, approximately 38% of students use the internet for at least four hours per weekday outside of school, while 51% of students use the internet for at least four hours per weekend day. In each case, females are slightly more likely than males to be using the internet for these prolonged periods per day.

In terms of ethnicity, one feature that stands out is that internet use outside of school for over six hours per day (on weekdays and weekends) is most predominant amongst Māori students: 27% of Māori students report using the internet outside of school on weekdays for over six hours per day, with this rate rising to 32% on weekends. We have no information on what types of material are being accessed by these prolonged internet users, and further research on the internet use of prolonged users is warranted.

NZES

Table 11 to Table 18 report descriptive statistics for different demographic groups and their internet access using 2017 NZES data. The definition of internet access for this dataset is whether an individual ticked yes to one or more of: having access to the internet at work, having access to the internet at home, having access to the internet on mobile or having access to the internet somewhere else, without ticking yes to having no access to the internet. The no internet category is applied to individuals who ticked no access to the internet and did not tick yes to any form of internet access. All data were weighted using the sample weights provided by NZES.

Total numbers in all PISA and NZES tables are after applying sampling weights.

Ethnicity and access to the internet

Table 11 shows the breakdown of access to the internet in 2017 by ethnicity. It shows that, of the identified ethnicities, Māori (12.23%) and Pasifika (10.55%) are the most likely not to have internet access.

Table 11: Ethnicity and access to the internet
Ethnicity No Internet Internet Unknown Total

European

8.89%

90.49%

0.61%

2,609

Māori

12.23%

87.13%

0.64%

336

Pasifika

10.55%

89.45%

0.00%

81

Asian

2.67%

97.33%

0.00%

199

Other

13.34%

86.66%

0.00%

18

No Response

10.43%

87.74%

1.83%

203

Total

9.00%

90.36%

0.64%

3,445

Gender and access to the internet

In Table 12 we see that there is little disparity between males and females. Those who identified as gender diverse (and those who did not identify their gender) have a much higher proportion without access to the internet (22.11% and 18.59% respectively) but the number of individuals in these groups is low (7 and 81 respectively).

Table 12: Gender and access to the internet
Gender No Internet Internet Unknown Total

Male

8.45%

91.10%

0.45%

1,636

Female

9.00%

90.27%

0.73%

1,721

Gender diverse

22.11%

77.89%

0.00%

7

Non response

18.59%

78.49%

2.92%

81

Total

9.00%

90.36%

0.64%

3,445

Age and access to the internet

Table 13 reports access to the internet broken down by age. Access to the internet decreases as the age group gets older, sharply so beyond age 65. By comparison with the youngest cohort (which has over 99% access) over 35% of those who are over 75 years old have no internet access.

Table 13: Age and access to the internet
Age Internet No internet Unknown Total

<26

99.17%

0.18%

0.65%

425

26-45

96.33%

3.67%

0.00%

1,053

46-65

92.33%

7.06%

0.61%

1,146

66-75

85.68%

13.88%

0.44%

487

>75

60.34%

36.54%

3.12%

334

Total

90.36%

9.00%

0.64%

3,445

Education level and access to the internet

Table 14 reports access to the internet, broken down by education level. It shows a trend that the higher qualification an individual has, the more likely it is that they will have internet access. Those with no qualification have the lowest proportion of individuals with access to the internet at 72.50%. This result may be partly explained by age since individuals in the oldest cohorts have fewer qualifications than those in younger cohorts. All groups with university level qualifications have at least 98% of individuals with access to the internet.

Table 14: Education level and access to the internet
Education No Internet Internet Unknown Total

No Qualification

25.61%

72.50%

1.89%

636

School Certificate/Level 1

14.36%

85.33%

0.31%

333

Sixth Form Certificate Level 2

1.99%

98.01%

0.00%

200

University Entrance

3.51%

96.49%

0.00%

130

Higher School Certificate

2.56%

97.44%

0.00%

73

University Entrance Bursary

15.43%

84.57%

0.00%

91

Bursary or School Level 3

0.09%

99.91%

0.00%

124

Another secondary qualification New Zealand

25.34%

74.45%

0.20%

55

Another secondary qualification overseas

2.64%

97.36%

0.00%

111

No Response

21.36%

76.84%

1.80%

131

National Certificate Level 4 post school

4.48%

94.55%

0.97%

357

Poly Tech

3.08%

96.26%

0.66%

333

Undergrad

0.31%

99.56%

0.14%

559

Masters Hons

0.67%

99.33%

0.00%

271

Doctorate

0.00%

98.94%

1.06%

41

Total

9.00%

90.36%

0.64%

3,445

Position inside or outside the workforce and access to the internet

In Table 15, we show access to the internet broken down by work situation. Those working full time (98.07%) and those studying at university or at another institution (98.86%) have the highest proportions of people with access to the internet. Surprisingly, those who are self-employed have a lower proportion with access to the internet (91.74%) than those who are employed full-time or part-time. Those who are retired (74.65%) and the disabled (71.17%) have the least access.

Table 15: Position inside or outside the workforce and access to the internet
Job No Internet Internet Unknown Total

Working full time

1.69%

98.07%

0.24%

1,568

Working part time

5.66%

94.16%

0.18%

555

Self-employed

7.83%

91.74%

0.44%

564

Unemployed

11.55%

88.38%

0.06%

146

Retired

24.19%

74.65%

1.17%

728

Disabled

27.30%

71.17%

1.52%

141

At school or university

0.66%

98.86%

0.48%

210

Working unpaid outside the home

6.37%

92.85%

0.77%

90

Working unpaid inside the home

4.59%

94.47%

0.94%

167

Total

8.47%

92.55%

0.51%

4,107

Area residing in and internet access

In Table 16, we report access to the internet broken down by the type of area in which an individual resides. Those living in a major city, defined as a place which has more than 100,000 people, have the highest proportion of people with access to the internet (92.73%). While it might be expected that those in a rural area or settlement would have the lowest proportion of people with access to the internet, it is those in larger country towns (10,000-25,000 population) who have the lowest proportion (87.44%).

Table 16: Area residing in and internet access
Area residing in Internet No Internet Unknown Total

Rural area or settlement (under 10,000 population)

90.39%

8.25%

1.36%

486

Country town (under 10,000 population)

88.59%

10.85%

0.56%

394

Larger country town (10,000-25,000 population)

87.44%

12.16%

0.40%

274

Large town (over 25,000 population)

90.46%

9.33%

0.22%

603

Major city (over 100,000 population)

92.73%

6.74%

0.53%

1,563

Total

91.03%

8.38%

0.60%

3,321

Housing situation and internet access

Access to the internet broken down by housing situation is reported in Table 17. Those who rent a house or flat from HNZC or the local (social housing) equivalent have the lowest proportion of people with internet access (69.36%). This is considerably lower than the other categories. By contrast, those who own a house or flat with a mortgage have the highest proportion with access to the internet (96.17%).

Table 17: Housing situation and internet access
Housing situation Internet No internet Unknown Total

Own a house or flat mortgage free

87.12%

12.12%

0.75%

1,144

Own a house or flat with a mortgage

96.17%

3.45%

0.38%

950

Rent a house privately as a family

94.07%

5.61%

0.31%

407

Rent a house or flat from HNZC or local equivalent

69.36%

30.15%

0.49%

124

Board or live in a hotel, hostel, rest home, or temporary

80.69%

15.17%

4.14%

92

Rent a house with a group of individuals

91.92%

8.08%

0.00%

170

Live with parents or other family members

94.17%

5.50%

0.33%

414

Total

90.88%

8.54%

0.58%

3,302

Voting, civic participation and internet access

Tables 18a, 18b and 18c show the engagement of people with and without internet access in terms of voting behaviour (in each of general and local government elections) and other forms of civic participation (defined as one or more of: signing a petition, making a select committee submission, making a royal committee submission or consulting with government).

Each of these activities can be considered as separate social capital outcomes. People with internet have a higher turnout in general (but not local) elections than people without internet, and are much more likely to engage in other forms of civic participation.

These results do not control for other characteristics of the individual. In section 6, we further investigate the relationship between internet access and these social capital outcomes, and also examine the relationship of internet access with generalised dissatisfaction of an individual.

Table 18a: General election 2017
  Internet No internet

Cast a vote

83.36%

73.36%

Chose not to cast a vote

8.64%

12.14%

Didn't manage to vote

5.74%

8.50%

No Response

2.25%

6.00%

Total

100%

100%

Table 18b: Local elections 2016
  Internet No internet

Voted

49.81%

51.27%

Did not vote

31.25%

27.90%

Don’t know or can’t remember

18.95%

20.83%

Total

100%

100%

Table 18c: Other civic participation
  Internet No internet

Yes

40.34%

19.21%

No

51.78%

62.29%

Unknown

7.88%

18.50%

Total

100%

100%

We provide extra information, based on the longitudinal NZES information for 2011, 2014 and 2017 in the Appendix. Table A14 reports the pattern of internet access for individuals who are observed in each wave of the longitudinal sample. It shows considerable persistence in internet access, so that once an individual gains access, they are very likely to retain access. A small portion of respondents lose access in one or more waves. More common is a rump of respondents who have never had internet access. This group comprises 8.5% of the sample (using 2017 sample weights).

Other Appendix tables detail internet access in 2011, 2014 and 2017 based on work situation, ethnicity and age. For ethnicity, we base the categories on the 2011 information (i.e. their ethnicity in 2011), while for work situation and age we report access based on their current status. (Note there are small numbers for some categories, and in these cases the proportions of people having internet access should be ignored.)

Notable points from the Appendix tables include part-time workers substantially increasing internet access from 2011 to 2014 (and further again in 2017). The work situation table shows an increase followed by a decrease in internet access of retirees over the period. When we examine the same issue by age, we find a steady increase in internet access for the 66-75 year age group, while the over 75 year age group records an increase and then a decrease in internet access over the two intervals (2011-2014 followed by 2014-2017).

NZCVS

Table 19 to Table 24 provide descriptive statistics for different demographic groups and their internet access using 2018 NZCVS data. As part of the response to the question: ‘In the last 12 months, has a computer or internet-enabled device belonging to you or anyone else living in your household, been infected or interfered with, for example by a virus or someone accessing it without permissions?’ there is a possible response of ‘Not applicable – Nobody in this household has owned a computer or Internet-enabled device in the last 12 months’. This category is used here to represent those who do not have access to the internet. The results for this question are shown for a variety of demographic groups. All data were randomly rounded to base three using rules defined by Statistics NZ.

Each of the NZCVS tables presents three categories (plus ‘don’t knows’). The third group in each table reports those having no internet access over the previous 12 months. This corresponds to our no internet access groups for NZES and PISA. The first group in each table is also of policy interest: this group reports having had an internet-related violation such as a virus infection or unauthorised access over the past year.

Ethnicity and internet issue

Table 19 shows responses broken down by ethnicity. The highest proportion of individuals who do not have access to the internet are those who identify as Pasifika (7.9%). By contrast Māori and Europeans report very similar internet access. Māori and Pasifika are the most likely to report an internet violation.

Table 19: Ethnicity and internet issue
Ethnicity Violation: Yes Violation: No No internet Don't Know Total

European

3.9%

90.2%

5.4%

0.6%

5,442

Māori

5.3%

89.2%

5.0%

0.5%

2,304

Pasifika

5.1%

86.9%

7.9%

0.2%

492

Asian

2.4%

95.8%

1.2%

0.7%

759

Other

1.6%

94.4%

2.4%

1.6%

126

Total

4.1%

90.4%

5.0%

0.6%

9,123

Note: In all NZCVS tables, no internet means nobody in the household has owned a computer or internet-enabled device in the last 12 months. To ensure confidentiality all cells are random rounded to base three using the rules defined by Stats NZ.

Sex and internet issue

In Table 20, we show responses broken down by sex. The results show no gender divide for internet access with males and females having the same proportion of individuals without access to the internet (5.4%). Males and females also show the same rate of internet violation over the year.

Table 20: Sex and internet issue
Sex Violation: Yes Violation: No No internet Don't Know Total

Female

3.9%

90.2%

5.4%

0.6%

4,608

Male

3.9%

90.2%

5.4%

0.6%

3,423

Total

3.9%

90.2%

5.4%

0.6%

8,031

Age and internet issue

Responses according to age are reported in Table 21. Consistent with NZES, as age increases, the proportion of those without access to the internet increases. For example, 0.9% of 15 to 19 year olds do not have access to the internet whereas 15.8% of those 65 years have no internet access. There is no clear trend in rates of internet violation by age, although there may be some need to ensure that the youngest users (with a comparatively high 5.1% violation rate) are made fully aware of the risks that may arise from internet use.

Table 21: Age and internet issue
Age groups Violation: Yes Violation: No No internet Don't Know Total

15-19 Years

5.1%

93.7%

0.9%

0.3%

333

20-29 Years

3.2%

94.4%

2.0%

0.4%

1,176

30-39 Years

4.0%

94.0%

1.3%

0.8%

1,410

40-49 Years

5.8%

91.1%

2.4%

0.7%

1,353

50-59 Years

4.1%

92.5%

3.0%

0.4%

1,407

60-64 Years

3.4%

89.6%

6.1%

1.0%

624

65 Years and Over

2.4%

81.2%

15.8%

0.6%

1,725

Total

3.9%

90.2%

5.4%

0.7%

8,028

Household income and internet issue

Table 22 shows responses according to household income (similar results are found when using personal income in place of household income). As household income increases above $20,000, the proportion of people with access to the internet increases. The group with the highest proportion of people without access to the internet is the $10,001 to $20,000 a year group (21.1%). This group may include substantial numbers of single older New Zealanders as the single rate of New Zealand Superannuation is just below $20,000 p.a. It is also likely to include a substantial number of beneficiaries under the age of 65 (for instance, the annual net rate of unemployment benefit for a single parent with children is less than $15,000 p.a.).

Table 22: Household income and internet issue
Household income Violation: Yes Violation: No No internet Don't Know Total

$10,000 or less

4.5%

91.9%

2.8%

0.8%

357

$10,001-$20,000

4.4%

73.8%

21.1%

0.6%

474

$20,001-$30,000

3.4%

81.7%

14.4%

0.6%

1,017

$30,001-$40,000

3.2%

87.7%

8.5%

0.7%

744

$40,001-$50,000

4.3%

90.1%

5.2%

0.4%

699

$50,001-$60,000

4.0%

92.5%

2.7%

0.9%

702

$60,001-$70,000

4.0%

92.1%

3.2%

0.7%

696

$70,001-$100,000

3.2%

94.5%

1.6%

0.7%

1,299

$100,001-$150,000

3.9%

94.8%

0.8%

0.5%

1,230

$150,001 or more

4.8%

94.5%

0.4%

0.4%

810

Total

3.9%

90.3%

5.4%

0.6%

8,028

Employment status and internet issue

Table 23 reports responses according to employment status. The group with the highest proportion of people without access to the internet are those who are retired (17.3%) and those not actively seeking work (16.6%). These proportions are much higher than for those who are unemployed (5.2%). By contrast, only 1.0% of those studying and 1.9% of those employed do not have access to the internet. Those who are not actively seeking work are most at risk of an internet violation (7.3%) followed by people who are unemployed (5.7%). Consistent with the age results, people who are studying are also at heightened risk of internet violation (5.2%).

Table 23: Employment status and internet issue
Employment status Violation: Yes Violation: No No internet Don't Know Total

Employed

4.0%

93.7%

1.9%

0.5%

4,986

Unemployed

5.7%

88.0%

5.2%

1.1%

369

NILF - Retired

2.5%

79.5%

17.3%

0.8%

1,431

NILF - Home or caring duties

2.6%

91.6%

4.5%

1.3%

465

NILF - Studying

5.2%

93.9%

1.0%

0.0%

309

NILF - Not actively seeking work

7.3%

74.6%

16.6%

1.5%

204

Other (not specified)

4.5%

89.8%

5.3%

0.4%

246

Total

3.9%

90.2%

5.4%

0.6%

8,010

Disability status and internet issue

Table 24 shows responses broken down by disability status. There is a large difference between those who are disabled (17.2%) and those who are not disabled (4.7%) with regards to the proportion of individuals who do not have access to the internet. This is likely to be of policy concern since one might anticipate that disabled people (and especially physically disabled people) could benefit most from internet access. Furthermore, disabled people are more likely than others to have received an internet violation, emphasising their at-risk status with respect to secure internet access.

Table 24: Disability status and internet issue
Disability status Violation: Yes Violation: No No internet Don't Know Total

Disabled

6.7%

75.2%

17.2%

0.5%

447

Not disabled

3.7%

91.0%

4.7%

0.6%

7,581

Total

3.9%

90.2%

5.3%

0.6%

8,028

6. Results: Wellbeing and the Internet

In this section we report relationships between internet access, internet use and indicators of wellbeing. Each of the PISA, NZES and NZCVS surveys are used to shed light on these relationships.

PISA

Table 25a to Table 25g report the results of a regression of the subjective wellbeing (SWB) proxy variable against internet access, other ICT variables and demographic variables. The demographic variables which are included in the equation are not reported; they are available on request.

The coefficients should not be used to indicate causal relationships; rather they demonstrate associations of wellbeing with internet use and availability. The results indicate that students who have use of a desktop at home, laptop at home, tablet at home or wireless internet at school are better off in terms of wellbeing than are those who do not have these items. Similarly, those who access the internet outside of school on weekdays have higher wellbeing relative to those who do not. These results may reflect family circumstances rather than ICT access per se.

Potentially of more policy interest is the finding that the positive relationship between wellbeing and internet access outside of school on weekdays declines monotonically the more time an individual spends online during a week day outside of school (see Tables). No significant relationship is found between SWB and time spent on the internet outside of school on weekends or at school on weekdays.

The results relating to internet use at home on weekdays (for those who have internet access) suggest that while access to the internet may be beneficial, the beneficial association declines as students spend longer online (see Figure 1 below). Our results indicate no beneficial association once students are online for over two hours per weekday outside of school (and no benefits of use on weekends). These results are similar to the results relating to overall screen-time use for UK adolescents reported by Przybylski and Weinstein (2017).

Note: For internet use at home on weekdays, we do not reject the null hypothesis of equal coefficients for 1-30 minutes per day and 31-60 minutes per day (p=0.5824), 1-2 hours per day (p=0.1315) and unknown time per day (p=0.4922); we conclusively reject equal coefficients for 1-30 minutes per day and each of the longer usage periods (p=0.0202, p=0.0012 and p<0.001 respectively).

Our findings may suggest that students limit their time spent on the internet outside of school on weekdays to less than two hours. In this light, it is worth highlighting the actual time that students report spending on the internet each day. These results are shown in the Appendix Table A9. Approximately 15% of 15 year olds (including 27% of Māori students) report using the internet for more than 6 hours a day (on a weekday outside of school), while over half report more than two hours’ use.

Figure 1: Relationship between subjective wellbeing and time spent on internet outside of school on weekdays.

Figure 1: Relationship between subjective wellbeing and time spent on internet outside of school on weekdays.

Read detailed description of Figure 1

This bar graph shows how time spent on the internet per weekday contributes to subjective wellbeing. The graph suggests that that while access to the internet may be beneficial, the beneficial association declines as students spend longer online.

The relationship between wellbeing and internet access at home on weekdays

Time spent on internet at home weekdays (base category = no time spent)

Coefficient (standard error)

1-30 minutes per day

0.674** (0.199)

31-60 minutes per day

0.601** (0.194)

1-2 hours per day

0.472* (0.193)

2-4 hours per day

0.358 (0.194)

4-6 hours per day

0.217 (0.197)

Over 6 hours per day

0.128 (0.202)

Unknown

0.447 (0.356)

The relationship between wellbeing and internet and ICT access

In the following tables, demographic controls are included (but not reported) for ethnicity, gender, father’s education and mother’s education.

The number of observations used for the regression in the following tables was 4,001. To determine the goodness of fit, we used adjusted R^2. This determines the amount of variation in the wellbeing proxy explained by the variables included, running from 0 to 1. The adjusted R^2 found was 0.0787

As for all other regression tables * p<0.05; ** p<0.01.

Table 25a: The relationship between wellbeing and internet access at school
Time spent on internet at school (base category = no time spent) Coefficient (standard error)

1-30 minutes per day

-0.030 (0.075)

31-60 minutes per day

-0.014 (0.078)

1-2 hours per day

-0.084 (0.081)

2-4 hours per day

-0.046 (0.089)

4-6 hours per day

-0.087 (0.109)

Over 6 hours per day

-0.028 (0.123)

Unknown

0.030 (0.299)

Table 25b: The relationship between wellbeing and internet access at home on weekdays
Time spent on internet at home weekdays (base category = no time spent) Coefficient (standard error)

1-30 minutes per day

0.674** (0.199)

31-60 minutes per day

0.601** (0.194)

1-2 hours per day

0.472* (0.193)

2-4 hours per day

0.358 (0.194)

4-6 hours per day

0.217 (0.197)

Over 6 hours per day

0.128 (0.202)

Unknown

0.447 (0.356)

Table 25c: The relationship between wellbeing and internet access at home on weekends
Time spent on internet at home weekends (base category = no time spent) Coefficient (standard error)

1-30 minutes per day

0.109 (0.199)

31-60 minutes per day

0.242 (0.197)

1-2 hours per day

0.286 (0.195)

2-4 hours per day

0.292 (0.195)

4-6 hours per day

0.246 (0.197)

Over 6 hours per day

0.197 (0.199)

Unknown

0.382 (0.296)

Table 25d: The relationship between wellbeing and access to digital devices
Variable Coefficient (standard error)

Available desktop at home which is used

0.149** (0.040)

Available laptop at home which is used

0.231** (0.049)

Available tablet at home which is used

0.080* (0.041)

Available Internet connection at home which is used

0.069 (0.092)

Available cell phone at home with internet access which is used

-0.027 (0.042)

Available desktop at school which is used

0.052 (0.053)

Available laptop at school which is used

0.046 (0.041)

Available tablet at school which is used

-0.031 (0.048)

Available computer at school with internet which is used

-0.058 (0.063)

Available wireless internet connection at school which is used

0.198** (0.053)

Table 25e: The relationship between wellbeing and age of first access to digital devices
Variable Coefficient (standard error)

Age first used a digital device: 7-9 years old

-0.018 (0.053)

Age first used a digital device: 10-12 years old

-0.037 (0.060)

Age first used a digital device: 13 years old or older

-0.047 (0.090)

Have never used a digital device

-0.131 (0.335)

Unknown age first used a digital device

0.405 (0.236)

Table 25f: The relationship between wellbeing and age of first access to a computer
Variable Coefficient (standard error)

Age first used a computer: 7-9 years old

-0.020 (0.054)

Age first used a computer: 10-12 years old

-0.131 (0.078)

Age first used a computer: 13 years old or older

-0.050 (0.129)

Have never used a computer

0.194 (0.561)

Unknown age first used a computer

-0.144 (0.264)

Table 25g: The relationship between wellbeing and age of first access to the internet
Variable Coefficient (standard error)

Age first used the internet: 7-9 years old

-0.044 (0.059)

Age first used the internet: 10-12 years old

-0.016 (0.071)

Age first used the internet: 13 years old or older

-0.176 (0.116)

Have never used the internet

-0.746 (0.456)

Unknown age first used the internet

-0.174 (0.259)

NZES

The relationship between wellbeing and internet access is investigated using NZES based on a principal component dissatisfaction variable constructed using the process described in our Methodology section. We also examine the relationship between wellbeing and three outcome variables related to social capital: whether the respondent voted in the 2017 general election, whether they voted in the 2016 local body elections, and whether they have undertaken some other form of civic participation in the 12 months leading up to the 2017 survey (where civic participation includes signing a petition, making a select committee submission, making a royal committee submission and/or consulting with government).

On average, we find that people without internet access record higher dissatisfaction than do people with internet access (4.152 vs 3.802 respectively). A regression (without other added control variables) of the dissatisfaction variable on internet access shows a significant positive relationship between dissatisfaction and no internet access. However the lack of demographic and other controls means that this raw regression may not provide an accurate picture of the relationship between wellbeing and internet access.

In column 1 of Table 26 we show the results from an OLS regression of the dissatisfaction variable on internet access plus demographic variables. The coefficients indicate associations, and again we cannot draw causal conclusions from them. Demographics controlled for in this regression (but not reported) include gender, working situation, education level, age, income level and ethnicity.

The results indicate that the wellbeing (dissatisfaction) of those who are without internet (and those with unknown internet status) is not different to that of people with internet once we control for demographic and other factors. One possible reason for lack of a clear result here is that the proxy dissatisfaction variable may not be an adequate representation of (a lack of) wellbeing. However the (unreported) results for the demographic controls in the estimated equation are consistent with other SWB findings. For instance: those who identify as Māori are more dissatisfied than those who identify as European, and lower income groups are more dissatisfied than are the highest income groups. These results signal that the principal component variable is likely to be a reasonable measure of dissatisfaction.

Another possible reason for lack of a clear result is that we are testing the relationship between internet access and wellbeing (dissatisfaction). The PISA results suggest that this relationship may depend on the extent of internet access by users, with any positive association between access and wellbeing tailing off sharply as daily internet use increases. We have no measure of intensity of use in the NZES data so our result may reflect an offsetting combination of positive effects for low internet users and negative effects for higher internet users.

Columns 2, 3 and 4 of Table 26 investigate the relationship of internet access with the three social capital outcome variables. In each case, we estimate a logit regression testing whether (lack of) internet access is associated with the social capital outcome. Demographic variables (as above) are included but not reported. Results for the internet variables are reported as odds ratios: a coefficient less than one means that individuals in that category are less likely to engage in the social capital outcome than are other individuals. (The significance test relates to whether the coefficient is different from one.)

We find that those without internet access were (statistically significantly) less likely to vote in the 2017 general election and less likely to be involved in various other forms of civic participation. There is no significant relationship of internet access and voting in the 2016 local body elections. The general election and civic participation results are consistent with the idea that people who are not engaged with others through the internet are also less engaged via more formal aspects of social capital such as civic participation.

Relationship between internet access, wellbeing and social capital indicators

Variable information

Dissatisfaction is the principal component variable described in section 4; the Dissatisfaction regression is estimated using OLS.

GE Vote is a binary variable with 1= voted in 2017 general election (0 otherwise), estimated using logit regression.

LE Vote is a binary variable with 1= voted in 2016 local election (0 otherwise), estimated using logit regression.

Civic Part is a binary variable with 1= someone who signed a petition, made a select committee submission, made a royal committee submission and/or consulted with the government in the 12 months leading up to the 2017 survey (0 otherwise), estimated using logit regression.

Logit results are presented as odds ratios (so a coefficient >1 indicates a positive relationship) with pseudo as the measure of goodness of fit. In each case, unreported demographic variables are included for gender, age, work status, ethnicity, income, qualifications.

Note: Standard errors are in brackets.

Table 26: Relationship between internet access, wellbeing and social capital indicators
Variable Dissatisfaction GE Vote LE Vote Civic Part

No internet access

0.078 (0.074)

 

0.460* (0.142)

 

1.035 (0.202)

 

0.634** (0.102)

 

Internet access unknown

0.051 (0.229)

 

0.334 (0.265)

 

0.774 (0.420)

 

1.775 (0.843)

 

Number of observations

2,127

3,256

2,697

3,241

OLS: Adjusted R2

Logit: pseudo R2

0.1735

0.1411

0.0985

0.0709

NZCVS

Life satisfaction and internet issue

Table 27 reports NZCVS responses broken down by reported life satisfaction (which is measured on a 0 to 10 scale, similar to that used in New Zealand’s General Social Survey). Those with the lowest level of life satisfaction (0-6) have the highest likelihood of no internet access (8.0%). This mirrors the PISA result in which no internet access is associated with lower subjective wellbeing. The rate of internet access appears unrelated to wellbeing for those whose life satisfaction ranges from 7 to 9. Those reporting life satisfaction of 10 have slightly lower rates of internet access, but this association likely reflects age, with retired people generally having high life satisfaction and lower rates of internet access.

It is notable from Table 27 that the rate of internet violations decreases as life satisfaction increases. This suggests that people who are more vulnerable in general (i.e. who have lower life satisfaction) are also those who are more vulnerable to internet violations.

Table 27: Life satisfaction and internet issue
Life satisfaction Violation: Yes Violation: No No internet Don't Know Total

0-6

5.6%

85.3%

8.0%

1.1%

1,068

7

4.7%

89.7%

4.9%

0.7%

1,221

8

3.8%

91.1%

4.6%

0.6%

2,448

9

3.4%

91.6%

4.7%

0.3%

1,545

10

2.6%

91.3%

5.6%

0.6%

1,725

Total

3.8%

90.2%

5.3%

0.6%

8,007

7. Conclusions

Our findings relating to internet access and exclusion replicate a number of existing findings for New Zealand. In particular, we find that Pasifika, Māori, those living in larger country towns, and older members of society are comparatively less likely to have internet access.

With respect to older people, the gap in access is much greater for those aged over 75 years than for those aged between 65 and 74 years old. This may suggest that the lack of internet access for those over 75 years is more of a cohort than an age effect – i.e. as the current 65-74 year olds graduate to the over 75 year category, internet access rates may not reduce sharply in the way that is currently observed. Hence a watching brief may be all that is required to monitor internet access rates as people in their sixties continue to age.

We find also that people living in social housing, unemployed people, those not actively seeking work, and disabled individuals are more likely than others to lack internet access. Some people who are unemployed and some who are not actively seeking work may be disengaged from other aspects of society. They are also likely to suffer severe financial constraints, so their lack of internet access is not surprising. Their lack of access may also not be particularly amenable to policy intervention other than by helping these individuals find secure employment.

The very large gaps in internet access for those who live in social housing and for people who identify as being disabled are perhaps the most disturbing. However, these gaps are also potentially amenable to policy interventions.

Most social housing is owned by the state (Housing NZ) or by local authorities or NGOs. In each case, the social housing provider – as opposed to the tenants (who may well suffer from multiple forms of disadvantage) – could take the initiative to install WiFi (or other technologies) to enable internet access by tenants. In the modern age, provision of such infrastructure may be considered similar to provision of water, sewerage and electricity, and may be particularly valuable to marginalised tenants. Such provision may also help to address the lower rates of internet access for Pasifika (and Māori) students.

Many people who are disabled are already subject to some form of care by the authorities (e.g. DHBs) and/or NGOs that deal with specific forms of disability. These authorities may consider enabling internet use for their clients as a key intervention designed to improve the opportunities for disabled people to connect with the rest of society.

Our work examining the connections between internet access, internet use and wellbeing indicate certain associations between internet access and wellbeing – but these associations are not necessarily causal. The NZCVS data indicate that those who do not have internet access tend to have lower wellbeing than those who do have access. The NZES data reveal a similar relationship when we do not control for other factors but that relationship is not found to be present when we control for demographic and other factors. The NZES data do show, however, that those without internet access are less engaged in civic activities such as voting in general elections and in making submissions to government, even after we control for other factors.

Perhaps our most intriguing result with respect to wellbeing concerns the association between subjective wellbeing and internet use by adolescents outside of school on weekdays. This work, based on the PISA data for 15 year olds, controls for other demographic influences. We find that those without internet access tend to have lower wellbeing than those with access (a result which may reflect family circumstances). For those who do access the internet on weekdays outside of school, we find that increases in daily internet use are associated with decreases in levels of wellbeing. Once daily internet use (on weekdays outside of school) exceeds about two hours, we find no positive association of internet use and wellbeing.

Interpretation of this result requires caution because it could, for instance, be that those with poor wellbeing choose to lock themselves away from direct human contact and instead interact with the world through the internet. However, from an intuitive angle, the proportion of youth who report extended internet use on weekdays (outside of school) may well be of concern. We find that 15% of 15 year olds (including 27% of Māori students) report using the internet for more than 6 hours per day on a weekday outside of school, while over half report more than two hours’ use.

We recommend further investigation of the wellbeing effects of such extended use of the internet – both for adolescents and, if the data were available, for children and adults. As always, the issues of causality pose a problem for such research. However, scheduled new surveys offer avenues to address some of these issues. Three recent and scheduled surveys, in particular, offer promise since all three will be linked into Statistics New Zealand’s Integrated Data Infrastructure (IDI) that contains a wealth of information about surveyed individuals to which the relevant surveys can be linked.

The first of these surveys is the 2018 NZCVS – i.e. the survey used here. The responses to this survey are currently being linked into the IDI. Future NZCVS surveys will similarly be linked to the IDI. The second set of surveys are the 2018 and scheduled 2021 PISA surveys. The 2018 survey is expected to be linked into the IDI in early 2020. The third survey is the scheduled PIAAC second cycle survey to be conducted over 2021/22 (with results released in 2023).

By linking these survey results into other (especially prior) characteristics of the surveyed individuals, we will be able to control much better for personal and locational traits that affect both wellbeing and internet (and other ICT) use. While the internet survey data will be cross-sectional, it will be possible to use statistical techniques such as propensity score matching to compare the wellbeing outcomes for otherwise similar individuals who have different internet use and access. From an internet access stance, we will also be able to control much better for geographical and other characteristics when examining the types of people who do not have access. Doing so will enable consideration of more specific policies that are targeted at groups who may benefit from internet use but who do not currently access the internet.

Studies across a wide range of social policies in New Zealand have benefitted from analysis of specific survey data linked into the IDI. We expect that similar benefits will arise from a focused set of studies relating to the internet (and to broader ICT use) using the surveys outlined above.

References

Andrade, D., Hedges, M., Karimikia, H. and Techatassanasoontorn A. 2017. “World Internet Project New Zealand. The internet in New Zealand in 2017.” New Zealand Work Research Institute, Auckland.

Auckland University of Technology. 2018. “Current State of Broadband Usage of Rural Communities in New Zealand.” Rural Broadband Usage Survey Project Team, Auckland.

Bughin, J., Hazan, E., Allas, T., Hjartar, K., Manyika, J., Sjatil, P.E., Shigna, I. 2019. “Tech for Good Smoothing disruption, improving well-being”. McKinsey Global Institute. New York, United States.

Digital Inclusion Research Group. 2017. “Digital New Zealanders: The Pulse of our Nation.” MBIE and DIA, Wellington, New Zealand.

Evans, L., and Strickland, E. Eds. 2018 “Out of the Maze: Building digitally inclusive communities,” SensePublishers, Rotterdam, New Zealand.

French, T., Quinn, L., & Yates, S. 2018. “Digital Motivation : Exploring the reasons people are offline (PDF),” Good Things Foundation, UK.

Greenbrook-Held, J., Morrison, P.S. 2011. “The domestic divide: Access to the Internet in New ZealandNew Zealand Geographer. Volume 67. Pp. 25-38.

InternetNZ 2017 “Solving Digital Divides Together

MBIE 2015. “Māori me te Ao Hangarau 2015 The Māori ICT Report 2015 (PDF).” He kai kei aku ringa. The Crown- Māori Economic Growth Partnership, Wellington, New Zealand.

Ministry of Business, Innovation & Employment & Stats NZ. 2019. “Digital Nation Domain Plan 2019 (PDF)”. New Zealand Government. Wellington, New Zealand.

Przybylski, A., and Weinstein, N. 2017. “A Large-Scale Test of the Goldilocks Hypothesis: Quantifying the Relations Between Digital-Screen Use and the Mental Well-Being of Adolescents (PDF).” Psychological Science, 28(2), 204-215.

Smith, P., Bell, A., Miller, M., and Crothers, C. 2016. “World Internet Project New Zealand. New Zealand Internet Trends in New Zealand 2007 – 2015” Institute of Culture, Discourse & Communication, Auckland University of Technology.

Statistics New Zealand. 2017 “New Zealand internet going mobile

Statistics New Zealand. 2018. “Internet-service-provider-survey-2018.” Stats NZ, Wellington, New Zealand.

Statistics New Zealand. 2018. “Stocktake for the digital nation domain plan 2018.” New Zealand Government, Wellington, New Zealand.

Stephanie, D. 2018. “Digital Divide Experiences from the Chatham Islands (PDF),” in Proceedings of the 9th Annual CITRENZ Conference (2018). pp. 22–29.

Appendices

Appendix — PISA data

Internet and other digital questions

Question 1

Are any of these devices available for use at home?

  • Desktop Computer
  • Portable laptop, or notebook
  • Tablet computer
  • Internet connection
  • Cell phone with internet access

For each device the following options were available:

  • Yes, and I use it
  • Yes, but I don’t use it
  • No
Question 2

Are there any of these devices available for you to use at school?

  • Desktop computer
  • Portable laptop, or notebook
  • Tablet computer
  • Internet-connected school computers
  • Internet connection via wireless network

For each device the following options were available:

  • Yes, and I use it
  • Yes, but I don’t use it
  • No
Question 3

How old were you when you first used a digital device?

  • 6 years old or younger
  • 7-9 years old
  • 10-12 years old
  • 13 years old or older
  • I had never used a digital device until today
Question 4

How old were you when you first used a computer?

  • 6 years old or younger
  • 7-9 years old
  • 10-12 years old
  • 13 years old or older
  • I had never used a computer until today
Question 5

How old were you when you first accessed the Internet?

  • 6 years old or younger
  • 7-9 years old
  • 10-12 years old
  • 13 years old or older
  • I have never used the Internet
Question 6

During at typical weekday, for how long do you use the internet at school?

  • No time
  • 1-30 minutes per day
  • 31-60 minutes per day
  • Between 1 hour and 2 hours per day
  • Between 2 hours and 4 hours per day
  • Between 4 hours and 6 hours per day
  • More than 6 hours per day
Question 7

During a typical weekday, for how long do you use the internet outside school?

  • No time
  • 1-30 minutes per day
  • 31-60 minutes per day
  • Between 1 hour and 2 hours per day
  • Between 2 hours and 4 hours per day
  • Between 4 hours and 6 hours per day
  • More than 6 hours per day
Question 8

On a typical weekend day, for how long do you use the internet outside of school?

  • No time
  • 1-30 minutes per day
  • 31-60 minutes per day
  • Between 1 hour and 2 hours per day
  • Between 2 hours and 4 hours per day
  • Between 4 hours and 6 hours per day
  • More than 6 hours per day

Wellbeing indicators

Question 1

Overall, how satisfied are you with your life as a whole these days? [Available for UK; not NZ.]

Sliding bar ranging 0-10 from 0, not at all satisfied to 10, completely satisfied

Question 2

Home possessions which includes compiled variables for: wealth, cultural possessions, home education resources, ICT resources and how many books are in the respondents home

The wealth variable uses a combination of household possessions. These household possessions include:

  • a room of their own
  • a link to the internet
  • a dishwasher
  • 3 country specific wealth items.

And the number of the following in their home:

  • cell phones with internet access
  • computers
  • cars
  • rooms with a bath or shower
  • e-book readers.

The cultural possessions at home uses a combination of household possessions. These household possessions include:

  • classic literature
  • books of poetry
  • works of art
  • books on art, music, or design
  • musical instruments.

Home educational resources uses a combination of household possessions. These household possessions include:

  • a desk to study at
  • a quiet place to study
  • a computer you can use for school work
  • educational software
  • books to help you with school work
  • a dictionary
  • technical reference books.

The ICT resources variable includes:

  • educational software
  • a link to the internet
  • cell phones with internet access
  • computers
  • tablet computers
  • e-book readers.
Question 3

Sense of belonging which attempts to summarise a student’s sense of belonging to school using 6 trend items. The answering format is a four point Likert scale with the belonging categories “strongly agree”, “agree”, “disagree”, and “strongly disagree”. A higher score or weighted likelihood estimate is related to a higher sense of belonging at school. It is adjusted to have a mean of 0 and a standard deviation of 1.

The following items are included in this variable:

  • I feel like an outsider (or left out of things) at school
  • I make friends easily at school
  • I feel like I belong at school
  • I feel awkward and out of place at school
  • Other students seem to like me
  • I feel lonely at school.
Question 4

Value of cooperating uses an answering format which is a four point Likert scale with the belonging categories “strongly agree”, “agree”, “disagree”, and “strongly disagree”.

The value cooperation variable includes the following:

  • I prefer working as part of a team to working alone
  • I find that teams make better decisions than individuals
  • I find that teamwork raises my own efficiency
  • I enjoy cooperating with my peers.
Question 5

Parent’s emotional support uses an answering format which is a four point Likert scale with the belonging categories “strongly agree”, “agree”, “disagree”, and “strongly disagree”.

The parent’s emotional support variable includes the following:

  • My parents are interested in my school activities.
  • My parents support my educational efforts and achievements.
  • My parents support me when I am facing difficulties at school.
  • My parents encourage me to be confident.
Question 6

Instrumental motivation uses an answering format which is a four point Likert scale with the belonging categories “strongly agree”, “agree”, “disagree”, and “strongly disagree”.

The instrumental motivation question includes the following:

  • Making an effort in my school science subject(s) is worth it because this will help me in the work I want to do later on.
  • What I learn in my subject(s) is important for me because I need this for what I want to do later on.
  • Studying my subject(s) is worthwhile for me because what I learn will improve my career prospects.
  • Many things I learn in my subject(s) will help me to get a job.
Question 7

Test anxiety uses an answering format which is a four point Likert scale with the belonging categories “strongly agree”, “agree”, “disagree”, and “strongly disagree”.

The test anxiety question includes the following:

  • I often worry that it will be difficult for me taking a test.
  • I worry that I will get poor at school.
  • Even if I am well prepared for a test I feel very anxious
  • I get very tense when I study for a test
  • I get nervous when I don't know how to solve a task at school.

Demographic indicators

Question 1

Are you female or male?

  • Female
  • Male.
Question 2

Does your mother have any of the following qualifications?

  • ISCED level 6
  • ISCED level 5A
  • ISCED level 5B
  • ISCED level 4.
Question 3

Does your father have any of the following qualifications?

  • ISCED level 6
  • ISCED level 5A
  • ISCED level 5B
  • ISCED level 4.
Question 4

Ethnicity based question, respondents picked all options that encapsulated their ethnicity.

  • Māori
  • Pasifika
  • Asian
  • Pākehā
  • other.

United Kingdom SWB regression (PISA 2015 data)

Variable Coefficient (standard error)

Constant

7.151** (0.020)

Home possessions

0.080** (0.019)

Sense of belonging

0.595** (0.021)

Value cooperating

0.248** (0.018)

Parental support

0.566** (0.019)

Instrumental motivation

0.083** (0.018)

Test anxiety

-0.554** (0.019)

Number of observations: 12,606

Adjusted R2: 0.2760

Appendix — NZES data

Internet and other digital questions

Do you have access to the internet? (tick all that apply)

  • No access to the internet.
  • Access to the internet at work.
  • Access to the internet at home.
  • Access to the internet on a mobile device.
  • Access to the internet somewhere else.

Demographic indicators

Question 1

Respondent's gender:

  • Male.
  • Female.
  • Transsexual or transgender.
Question 2

Age in years:

Scale response.

Question 3

Working full-time for pay or other income?

  • Yes.
  • No.
Question 4

Working part-time for pay or other income?

  • Yes.
  • No.
Question 5

Unemployed, laid off, looking for work?

  • Yes.
  • No.
Question 6

Retired?

  • Yes.
  • No.
Question 7

Disabled, unable to work?

  • Yes.
  • No.
Question 8

At school, university, or other educational institution?

  • Yes.
  • No.
Question 9

Working unpaid outside the home?

  • Yes.
  • No.
Question 10

Working unpaid within the home?

  • Yes
  • No
Question 11

For whom do you work, or did you last work if currently unemployed?

  • I am/was self-employed
  • a private company or business
  • a state or public agency or enterprise, central or local
  • a mixed public/private, or non-profit organisation
  • I have never been in paid employment
  • don’t know.
Question 12

Main ethnicity:

  • European
  • Maori
  • Pasifika
  • Asian
  • Kiwi or New Zealand
  • other.
Question 13

Household income between 1 April 2016 and 31 March 2017:

  • no income
  • $23,800 or Less
  • $23,801-$35,699
  • $35,700-$62,199
  • $62,200-$76,999
  • $77,000-$93,599
  • $93,600-$136,599
  • $136,600-$180,199
  • $180,200 or over
  • don’t know.
Question 14

Highest qualification:

  • no qualification
  • Level 1
  • Level 2
  • University Entrance
  • Higher School Certificate
  • University Entrance Bursary
  • Bursary or School Level
  • Another secondary qualification in New Zealand
  • Another secondary qualification overseas
  • no response
  • National Certificate Level 4 post school
  • Polytechnic
  • Undergraduate
  • Masters/Honours
  • Doctorate.
Question 15

Did you vote or not vote?

  • cast a vote
  • chose not to vote
  • didn’t manage to vote
  • no response.
Question 16

Size of area usually lived in:

  • A rural area or settlement (under 100,000 population).
  • A country town (under 10,000 population).
  • A larger country town (10,000 to 25,000 population).
  • A large town (over 25,000 population).
  • A major city (over 100,000 population).
Question 17

What is your current housing status?

  • Own a house or flat mortgage free.
  • Own a house or flat with a mortgage.
  • Rent a house privately as a family.
  • Rent a house or flat from HNZC or local equivalent.
  • Board or live in a hotel, hostel, rest home or temporary housing.
  • Rent a house with a group of individuals.
  • Live with parents or other family members.
Question 18

Did you vote in the most recent local elections?

  • Yes.
  • No.
  • Don’t know.
Question 19

Have you signed a petition (hard copy, not online)?

  • Have done in last five years.
  • Have not done in last five years, but did consider doing.
  • Have not done in last five years and did not consider doing.
  • Don’t know.
Question 20

Have you made a select committee submission?

  • Have done in last five years.
  • Have not done in last five years, but did consider doing.
  • Have not done in last five years and did not consider doing.
  • Don’t know.
Question 21

Have you made a royal committee submission?

  • Have done in last five years.
  • Have not done in last five years, but did consider doing.
  • Have not done in last five years and did not consider doing.
  • Don’t know.
Question 22

Have you taken part in consultation with the government?

  • Have done in last five years.
  • Have not done in last five years, but did consider doing.
  • Have not done in last five years and did not consider doing.
  • Don’t know.
Question 23

If NZ Māori are you registered with your iwi organisation?

  • Yes.
  • No.
  • Don’t know.

Questions used in the principal component analysis for NZES

Question 1

Does it make a difference who is in power?

  • It doesn’t make any difference who is in power.
  • It makes very little difference who is in power.
  • It makes some difference who is in power
  • It makes a reasonable amount of difference who is in power.
  • It makes a big difference who is in power.
  • Don’t know.
Question 2

Does voting make any difference to what happens?

  • Voting won’t make any difference to what happens.
  • Voting won’t make much difference to what happens.
  • Voting can make some difference to what happens.
  • Voting can make a reasonable amount of difference to what happens.
  • Voting can make a big difference to what happens.
  • Don’t know
Question 3

How satisfied with how democracy works in NZ?

  • Very satisfied.
  • Fairly satisfied.
  • Not very satisfied.
  • Not at all satisfied.
  • Don’t know.
Question 4

How good a job has the government done over the last three years?

  • A very good job.
  • A fairly good job.
  • A fairly bad job.
  • A very bad job.
  • Don’t know.
Question 5

In the last 12 months has the state of the NZ economy got better or worse?

  • Got a lot better.
  • Got a little better.
  • Stayed about the same.
  • Got a little worse.
  • Got a lot worse.
  • Don’t know.
Question 6

I don't think politicians and public servants care what people like me think.

  • Strongly agree.
  • Agree.
  • Neither.
  • Disagree.
  • Strongly disagree.
  • Don’t know.
Question 7

My vote really counts in elections.

  • Strongly agree.
  • Agree.
  • Neither.
  • Disagree.
  • Strongly disagree.
  • Don’t know.
Question 8

How likely is it that your household's income could be severely reduced in.

  • Very likely.
  • Somewhat likely.
  • Somewhat unlikely.
  • Very unlikely.
  • Don’t know.
Question 9

How widespread or unusual is corruption among politicians and public servants.

  • Very widespread.
  • Quite widespread.
  • Quite unusual.
  • Very unusual.
  • Don’t know.
Question 10

How likely is it that your household’s income could be severely reduced in 12 months?

  • Very likely.
  • Somewhat likely.
  • Somewhat unlikely.
  • Very unlikely.
  • Don’t know.
Question 11

In the last 12 months has the state of the NZ economy got better or worse?

  • Got a lot better.
  • Got a little better.
  • Stayed about the same.
  • Got a little worse.
  • Got a lot worse.
  • Don’t know.
Question 12

Over the next ten years how likely are you to improve your standard of living?

  • Very likely.
  • Somewhat likely.
  • Very unlikely.
  • Don’t know.

Appendix — NZCVS data

Internet question

In the last 12 months, has a computer or Internet-enabled device belonging to you or anyone else living in your household, been infected or interfered with, for example by a virus or someone accessing it without permissions?

  • Yes – how many times?
  • No.
  • Not applicable – Nobody in this household has owned a computer or Internet enabled device in the last 12 months.
  • Don’t know.

Wellbeing questions

Life satisfaction:

0 (least satisfied) to 10 (most satisfied).

Demographic indicators

Question 1

Ethnicity:

  • European.
  • Māori.
  • Pasifika.
  • Asian.
  • other.
Question 2

Sex:

  • Female.
  • Male.
Question 3

Age groups:

  • 15 to 19 years.
  • 20 to 29 years.
  • 30 to 39 years.
  • 40 to 49 years.
  • 50 to 59 years.
  • 60 to 64 years.
  • 65 years and over.
Question 4

Personal income:

  • $10,000 or less.
  • $10,001 to $20,000.
  • $20,001 to $30,000.
  • $30,001 to $40,000.
  • $40,001 to $50,000.
  • $50,001 to $60,000.
  • $60,001 to $70,000.
  • $70,001 to $100,000.
  • $100,001 or more.
Question 5

Household income:

  • $10,000 or less.
  • $10,001 to $20,000.
  • $20,001 to $30,000.
  • $30,001 to $40,000.
  • $40,001 to $50,000.
  • $50,001 to $60,000.
  • $60,001 to $70,000.
  • $70,001 to $100,000.
  • $100,001 to $150,000.
  • $150,000 or more.
Question 6

Employment status:

  • Employed.
  • Unemployed.
  • Not in Labour Force (NILF) – retired.
  • NILF-Home or caring duties.
  • NILF-Studying.
  • NILF-Not actively seeking work.
  • Other (not specified).
Question 7

Disability status:

  • Disabled.
  • Not disabled.

Appendix — PIAAC data

The PIAAC survey collects information on cognitive and workplace skills needed within the adult population in over 40 countries.

It is administered every ten years, with the latest information being collected from April 2014 to March 2015.

The survey had a sample size of 6,177 and was administered via a computer with an option for it to be administered by pen and paper.

There are questions on demographics and computer access inside and outside of work which we report here.

We show whether respondents used a computer inside or outside work broken down by demographic group.

The demographic groups include: gender, education level, working status and whether they were self-employed or not; in each case sampling weights are used.

Internet and other digital questions

Question 1

Have you ever used a computer? This includes mobile phones and other hand-held electronic devices that are used to connect to the Internet, check emails etc.

  • Yes.
  • No.
Question 2

Do you use a computer in your everyday life now outside work?

  • Yes.
  • No.
Question 3

Has the respondent experience with computer?

  • Yes.
  • No.

Wellbeing indicators

Question 1

Is the respondent male or female?

  • Male.
  • Female.
Question 2

Which of the qualifications on this card is the highest you have obtained?

  • No formal qualification or below ISCED 1.
  • ISCED 1.
  • ISCED 2.
  • ISCED 3C shorter than 2 years.
  • ISCED 3C 2 years or more.
  • ISCED 3A-B.
  • ISCED 3 (without distinction A-B-C, 2y+).
  • ISCED 4C.
  • ISCED 4A-B.
  • ISCED 4 (without distinction A-B-C).
  • ISCED 5B.
  • ISCED 5A, bachelor degree.
  • ISCED 5A, master degree.
  • ISCED 6.
  • Foreign qualification.
Question 3

Please look at this card and tell me which ONE of the statements best describes your current situation. If more than one statement applies to you, please indicate the statement that best describes how you see yourself.

  • Full-time employed (self-employed, employee).
  • Part-time employed (self-employed, employee).
  • Unemployed.
  • Pupil, student.
  • Apprentice, internship.
  • In retirement or early retirement.
  • Permanently disabled.
  • In compulsory military or community service.
  • Fulfilling domestic tasks or looking after children/family.
  • Other.
Question 4

In this job are you working as an employee or are you self-employed?

  • Employee.
  • Self-employed.

PIACC results

Note: In all PIAAC tables, total numbers are after applying sampling weights.

Table A1: Gender and computer use

Use computer inside or outside work
Response Male Female Total

Yes

88.1%

90.54%

89.36%

No

11.89%

9.46%

10.64%

Use computer outside work
Response Male Female Total

Yes

85.78%

88.99%

87.44%

No

14.15%

10.99%

12.51%

Not stated

0.07%

0.02%

0.05%

Experience using a computer
Response Male Female Total

Yes

96.96%

97.98%

97.49%

No

2.90%

1.88%

2.38%

Not stated

0.14%

0.14%

0.14%

Totals

  • Male: 1,332,091
  • Female: 1,417,628
  • Total: 2,749,719

Table A2: Highest educational attainment and use of computers in everyday life outside work

Table A2: Highest educational attainment and use of computers in everyday life outside work
  Use computer Don't use computer Not stated Total

ISCED 1

47.70%

52.30%

0.00%

49,168

ISCED 2

69.38%

30.62%

0.00%

284,050

ISCED shorter than 2 years

81.48%

18.52%

0.00%

248,009

ISCED 3C 2 years or more

85.12%

14.73%

0.15%

202,712

ISCED 3A-B

92.14%

7.86%

0.00%

406,932

ISCED 4C

87.07%

12.93%

0.00%

236,709

ISCED 5B

89.29%

10.71%

0.00%

363,230

ISCED 5A, bachelor degree

96.86%

2.96%

0.19%

493,265

ISCED 5A, master degree

97.76%

2.24%

0.00%

210,974

ISCED 6

100.00%

0.00%

0.00%

25,356

Foreign qualification

84.58%

15.42%

0.00%

175,767

Don't know

31.10%

68.90%

0.00%

1,653

Total

87.44%

12.51%

0.05%

2,697,826

Table A3: Highest educational attainment and the use of computers inside and outside work

Table A3: Highest educational attainment and the use of computers inside and outside work
  Use computer Don't use computer Total

ISCED 1

50.35%

49.65%

49,168

ISCED 2

75.04%

24.96%

284,050

ISCED shorter than 2 years

85.91%

14.09%

248,009

ISCED 3C 2 years or more

89.05%

10.95%

202,712

ISCED 3A-B

95.26%

4.74%

406,932

ISCED 4C

92.00%

8.00%

236,709

ISCED 5B

93.84%

6.16%

363,230

ISCED 5A, bachelor degree

99.03%

0.97%

493,265

ISCED 5A, master degree

99.62%

0.38%

210,974

ISCED 6

100.00%

0.00%

25,356

Foreign qualification

88.00%

12.00%

175,767

Don't know

49.58%

50.42%

1,653

Total

91.08%

8.92%

2,697,826

Table A4: Highest educational attainment and computer experience

Table A4: Highest educational attainment and computer experience
  Computer experience No computer experience Not stated Total

ISCED 1

76.35%

23.65%

0.00%

49,168

ISCED 2

93.42%

6.51%

0.07%

284,050

ISCED shorter than 2 years

97.26%

2.62%

0.12%

248,009

ISCED 3C 2 years or more

98.46%

1.39%

0.15%

202,712

ISCED 3A-B

99.10%

0.90%

0.00%

406,932

ISCED 4C

97.69%

2.31%

0.00%

236,709

ISCED 5B

98.39%

1.43%

0.18%

363,230

ISCED 5A, bachelor degree

99.47%

0.06%

0.47%

493,265

ISCED 5A, master degree

100.00%

0.00%

0.00%

210,974

ISCED 6

100.00%

0.00%

0.00%

25,356

Foreign qualification

94.76%

5.24%

0.00%

175,767

Don't know

49.58%

50.42%

0.00%

1,653

Total

97.49%

2.38%

0.14%

2,697,826

Table A5: Current work situation and the use of computers in everyday life outside work

Table A5: Current work situation and the use of computers in everyday life outside work
  Use computer Don't use computer Not stated Total

Full time employed (self-employed, employee)

89.11%

10.86%

0.03%

1,418,637

Part-time employed (self-employed, employee)

89.02%

10.82%

0.16%

476,979

Self-employed

89.55%

10.18%

0.27%

345,128

Unemployed

76.64%

23.36%

0.00%

144,921

Pupil, student

96.99%

3.01%

0.00%

258,947

Apprentice, internship

90.25%

9.75%

0.00%

10,405

In retirement or early retirement

81.84%

18.16%

0.00%

54,956

Permanently disabled

51.13%

48.87%

0.00%

51,690

Fulfilling domestic tasks or looking after children/family

80.87%

19.13%

0.00%

233,094

Other

80.55%

19.45%

0.00%

48,195

Total

87.68%

12.25%

0.07%

3,042,954

Table A6: Current work situation and the use of computers inside and outside work

Table A6: Current work situation and the use of computers inside and outside work
  Use computer Don't use computer Total

Full time employed (self-employed, employee)

94.16%

5.84%

1,418,637

Part-time employed (self-employed, employee)

92.92%

7.08%

476,979

Self-employed

94.86%

5.14%

345,128

Unemployed

79.21%

20.79%

144,921

Pupil, student

97.18%

2.82%

258,947

Apprentice, internship

90.25%

9.75%

10,405

In retirement or early retirement

82.71%

17.29%

54,956

Permanently disabled

51.13%

48.87%

51,690

Fulfilling domestic tasks or looking after children/family

81.94%

18.06%

233,094

Other

82.15%

17.85%

48,195

Total

91.51%

8.49%

3,042,953

Table A7: Current work situation and computer experience

Table A7: Current work situation and computer experience
  Computer experience No computer experience Not stated Total

Full time employed (self-employed, employee)

98.19%

1.69%

0.13%

1,418,637

Part-time employed (self-employed, employee)

97.82%

1.92%

0.26%

476,979

Self-employed

97.73%

2.00%

0.27%

345,128

Unemployed

94.84%

5.16%

0.00%

144,921

Pupil, student

99.88%

0.12%

0.00%

258,947

Apprentice, internship

100.00%

0.00%

0.00%

10,405

In retirement or early retirement

91.54%

8.46%

0.00%

54,956

Permanently disabled

81.91%

17.52%

0.57%

51,690

Fulfilling domestic tasks or looking after children/family

96.37%

3.44%

0.19%

233,094

Other

96.98%

3.02%

0.00%

48,195

Total

97.51%

2.33%

0.15%

3,042,953

Appendix — PISA: Additional tables

Table A8: Male and female internet use at school

Table A8: Male and female internet use at school

Time spent on the Internet per day

Female Male Total

No time

5.61%

10.08%

1,002

1 min to 30 min

23.88%

24.17%

1,767

31 min to 1 hour

19.47%

19.60%

3,350

1 to 2 hours

17.57%

14.13%

8,330

2 to 4 hours

10.45%

8.52%

14,068

4 to 6 hours

5.10%

3.85%

9,607

Over 6 hours

3.54%

3.64%

7,956

No Response

14.37%

16.01%

8,194

Total

100%

100%

54,274

Table A9: Male and female internet use outside of school on weekdays

Table A9: Male and female internet use outside of school on weekdays
Time spent on the Internet per day Female Male Total

No time

1.92%

1.77%

1,002

1 min to 30 min

2.97%

3.53%

1,767

31 min to 1 hour

5.35%

6.99%

3,350

1 to 2 hours

15.20%

15.5%

8,330

2 to 4 hours

27.11%

24.75%

14,068

4 to 6 hours

18.54%

16.87%

9,607

Over 6 hours

14.61%

14.71%

7,956

No Response

14.31%

15.87%

8,194

Total

100%

100%

54,274

Table A10: Male and female internet use outside of school on weekends

Table A10: Male and female internet use outside of school on weekends
Time spent on the Internet per day Female Male Total

No time

1.94%

1.65%

1,002

1 min to 30 min

2.77%

3.52%

1,767

31 min to 1 hour

4.23%

5.92%

3,350

1 to 2 hours

10.21%

10.11%

8,330

2 to 4 hours

21.09%

21.17%

14,068

4 to 6 hours

21.47%

17.12%

9,607

Over 6 hours

23.42%

24.15%

7,956

No Response

14.87%

16.37%

8,194

Total

100%

100%

54,274

Table A11: Ethnicity and internet use at school

Time spent on Internet per day (Māori, Pasifika, Pākehā, Māori and Pākeha)
Time spent on the Internet per day Māori Pasifika Pākehā Māori and Pākehā

No time

9.83%

10.16%

7.78%

7.97%

1 min to 30 min

24.48%

17.60%

25.70%

21.58%

31 min to 1 hour

13.40%

13.68%

22.12%

22.73%

1 to 2 hours

13.88%

12.53%

17.53%

15.89%

2 to 4 hours

5.76%

7.35%

10.49%

10.65%

4 to 6 hours

9.17%

4.01%

4.20%

5.31%

Over 6 hours

8.43%

3.54%

2.88%

3.96%

No Response

15.05%

31.13%

9.29%

11.92%

Total

100%

100%

100%

100%

Time spent on Internet per day (Asian, other, all other combinations, unknown)
Time spent on the Internet per day Asian Other All other combinations Unknown

No time

8.42%

4.06%

6.35%

1.71%

1 min to 30 min

29.46%

11.96%

21.03%

2.23%

31 min to 1 hour

16.69%

25.07%

19.06%

1.04%

1 to 2 hours

14.11%

13.57%

16.46%

1.85%

2 to 4 hours

9.67%

12.94%

8.20%

3.18%

4 to 6 hours

3.65%

2.45%

4.22%

0.00%

Over 6 hours

2.98%

3.44%

5.23%

0.00%

No Response

15.03%

26.52%

19.45%

89.98%

Total

100%

100%

100%

100%

Time spent on Internet per day (totals)
Time spent on the Internet per day Total

No time

4,266

1 min to 30 min

13,040

31 min to 1 hour

10,602

1 to 2 hours

8,596

2 to 4 hours

5,146

4 to 6 hours

2,427

Over 6 hours

1,950

No Response

8,248

Total

54,274

Table A12: Ethnicity and internet use outside school on weekdays

Time spent on Internet per day (Māori, Pasifika, Pākehā, Māori and Pākeha)
Time spent on the Internet per day Māori Pasifika Pākehā Māori and Pākehā

No time

1.66%

3.34%

1.45%

3.20%

1 min to 30 min

5.37%

3.43%

3.59%

2.79%

31 min to 1 hour

5.20%

7.52%

6.70%

6.16%

1 to 2 hours

8.33%

9.79%

18.75%

15.09%

2 to 4 hours

21.31%

13.80%

29.37%

22.97%

4 to 6 hours

15.52%

12.90%

19.18%

21.44%

Over 6 hours

26.88%

18.64%

11.73%

16.75%

No Response

15.72%

30.58%

9.22%

11.60%

Total

100%

100%

100%

100%

Time spent on Internet per day (Asian, other, all other combinations, unknown)
Time spent on the Internet per day Asian Other All other combinations Unknown

No time

1.09%

1.32%

3.70%

0.00%

1 min to 30 min

2.32%

1.26%

2.14%

0.00%

31 min to 1 hour

5.43%

4.47%

5.47%

1.26%

1 to 2 hours

13.79%

11.00%

11.29%

3.82%

2 to 4 hours

27.61%

27.91%

25.15%

1.83%

4 to 6 hours

17.67%

15.32%

15.65%

1.94%

Over 6 hours

16.80%

15.74%

17.40%

1.81%

No Response

15.29%

22.98%

19.20%

89.35%

Total

100%

100%

100%

100%

Time spent on Internet per day (totals)
Time spent on the Internet per day Total

No time

1,002

1 min to 30 min

1,767

31 min to 1 hour

3,350

1 to 2 hours

8,330

2 to 4 hours

14,068

4 to 6 hours

9,607

Over 6 hours

7,956

No Response

8,194

Total

54,274

Table A13: Ethnicity and internet use outside school on weekends

Time spent on Internet per day (Māori, Pasifika, Pākehā, Māori and Pākeha)
Time spent on the Internet per day Māori Pasifika Pākehā Māori and Pākeha

No time

1.70%

3.97%

1.41%

2.72%

1 min to 30 min

4.85%

3.11%

3.13%

4.08%

31 min to 1 hour

5.21%

6.37%

5.98%

3.19%

1 to 2 hours

6.59%

8.72%

11.86%

13.46%

2 to 4 hours

17.51%

10.33%

24.74%

18.76%

4 to 6 hours

16.12%

14.99%

20.40%

20.45%

Over 6 hours

31.89%

21.25%

22.87%

25.64%

No Response

16.12%

31.26%

9.62%

11.70%

Total

100%

100%

100%

100%
Time spent on Internet per day (Asian, other, all other combinations, unknown)
Time spent on the Internet per day Asian Other All other combinations Unknown

No time

1.08%

1.32%

3.36%

0.00%

1 min to 30 min

3.05%

0.00%

2.41%

0.00%

31 min to 1 hour

3.03%

1.26%

4.81%

1.04%

1 to 2 hours

6.24%

12.81%

8.02%

2.77%

2 to 4 hours

20.99%

22.84%

18.08%

1.83%

4 to 6 hours

22.20%

20.04%

18.51%

1.16%

Over 6 hours

27.31%

17.23%

24.88%

1.81%

No Response

16.10%

24.49%

19.93%

91.39%

Total

100%

100%

100%

100%

Time spent on Internet per day (totals)
Time spent on the Internet per day Total

No time

974

1 min to 30 min

1,708

31 min to 1 hour

2,757

1 to 2 hours

5,513

2 to 4 hours

11,467

4 to 6 hours

10,465

Over 6 hours

12,911

No Response

8,479

Total

54,274

Appendix — NZES: Longitudinal Tables

Table A14: Internet in 2011, 2014 and 2017 using different sampling weights

Note: Total numbers are after applying sampling weights. N and Y refer to no internet and internet for each year. For example, N, N, N refers to those who have no internet in 2011, 2014 and 2017 and N, Y, N refers to those who have no internet in 2011 and 2017 but have internet in 2014.

Table A14: Internet in 2011, 2014 and 2017 using different sampling weights (percentages represent the percentage of total sample)
  Using 2011 sample weights Using 2014 sample weights Using 2017 sample weights

Y,Y,Y

85.65%

82.52%

79.21%

N,N,N

5.29%

5.83%

8.51%

N,N,Y

1.70%

1.01%

1.91%

N,Y,N

1.65%

1.81%

2.87%

Y,N,N

0.32%

0.71%

1.37%

Y,Y,N

1.56%

2.05%

1.74%

Y,N,Y

0.25%

0.27%

0.31%

N,Y,Y

2.33%

4.33%

2.68%

Unknown

2.95%

2.47%

3.31%

Total

390

479

450

Table A15: Internet access in 2011, 2014 and 2017 and work in 2011, 2014 and 2017

Notes: Total numbers are after applying sampling weights. Percentages represent the percentage of demographic group.

Working full time
Year Internet No internet Unknown Total

2011

96.64%

3.36%

0.00%

204

2014

95.08%

4.72%

0.21%

231

2017

96.38%

2.24%

1.38%

192

Working part time
Year Internet No internet Unknown Total

2011

87.68%

12.32%

0.00%

68

2014

91.37%

5.56%

3.08%

96

2017

96.37%

3.63%

0.00%

69

Unemployed
Year Internet No internet Unknown Total

2011

89.58%

10.42%

0.00%

6

2014

78.03%

21.97%

0.00%

6

2017

79.32%

20.68%

0.00%

7

Disabled
Year Internet No internet Unknown Total

2011

76.69%

23.31%

0.00%

15

2014

96.11%

2.88%

1.01%

24

2017

58.73%

40.94%

0.33%

28

At school, university or other
Year Internet No internet Unknown Total

2011

94.31%

5.69%

0.00%

16

2014

100.00%

0.00%

0.00%

8

2017

100.00%

0.00%

0.00%

6

Retired
Year Internet No internet Unknown Total

2011

63.65%

36.35%

0.00%

69

2014

76.43%

22.26%

1.31%

85

2017

71.05%

28.71%

0.25%

143

Unpaid inside the home
Year Internet No internet Unknown Total

2011

97.77%

2.23%

0.00%

27

2014

95.63%

4.37%

0.00%

32

2017

92.56%

7.44%

0.00%

21

Unpaid outside the home
Year Internet No internet Unknown Total

2011

95.62

4.38%

0.00%

8

2014

93.86%

6.14%

0.00%

14

2017

89.22%

10.78%

0.00%

10

Self-employed
Year Internet No internet Unknown Total

2011

91.01%

8.99%

0.00%

84

2014

89.13%

10.47%

0.40%

104

2017

91.39%

8.61%

0.00%

96

Total
Year Internet No internet Unknown Total

2011

89.18%

10.82%

0.00%

497

2014

90.74%

8.40%

0.87%

600

2017

86.92%

12.54%

0.54%

573

Table A16: Internet access in 2011, 2014 and 2017 and 2011 ethnicity

Notes: Total numbers are after applying sampling weights. percentages represent the percentage of each ethnicity group.

European
Year Internet No internet Unknown Total

2011

88.61%

11.39%

0.00%

304

2014

92.32%

7.38%

0.30%

304

2017

89.13%

9.98%

0.89%

304

Māori
Year Internet No internet Unknown Total

2011

78.46%

21.54%

0.00%

21

2014

83.26%

16.53%

0.22%

21

2017

80.94%

18.59%

0.46%

21

Pasifika
Year Internet No internet Unknown Total

2011

34.91%

65.09%

0.00%

5

2014

36.52%

63.48%

0.00%

5

2017

92.89%

7.11%

0.00%

5

Asian
Year Internet No internet Unknown Total

2011

100.00%

0.00%

0.00%

14

2014

92.33%

0.00%

7.67%

14

2017

100.00%

0.00%

0.00%

14

Other
Year Internet No internet Unknown Total

2011

100.00%

0.00%

0.00%

14

2014

100.00%

0.00%

0.00%

14

2017

100.00%

0.00%

0.00%

14

European and Māori
Year Internet No internet Unknown Total

2011

95.34%

4.66%

0.00%

16

2014

95.34%

4.20%

0.47%

16

2017

99.32%

0.68%

0.00%

16

All other combinations
Year Internet No internet Unknown Total

2011

96.92%

3.08%

0.00%

5

2014

97.77%

2.23%

0.00%

5

2017

97.77%

2.23%

0.00%

5

No response
Year Internet No internet Unknown Total

2011

96.20%

3.80%

0.00%

11

2014

100.00%

0.00%

0.00%

11

2017

99.50%

0.50%

0.00%

11

Total
Year Internet No internet Unknown Total

2011

88.87%

11.13%

0.00%

 

390

2014

91.88%

7.59%

0.53%

390

2017

90.35%

8.93%

0.72%

390

Table A17: Internet access in 2011, 2014 and 2017 and age in 2011, 2014 and 2017

Notes: Total numbers are after applying sampling weights. Percentages represent the percentage of each age group.

<26
Year Internet No internet Unknown Total

2011

94.58%

5.42%

0.00%

16

2014

100.00%

0.00%

0.00%

12

2017

100.00%

0.00%

0.00%

2

26-45
Year Internet No internet Unknown Total

2011

96.56%

3.44%

0.00%

129

2014

97.71%

0.00%

2.29%

129

2017

89.62%

10.38%

0.00%

92

46-65
Year Internet No internet Unknown Total

2011

93.58%

6.42%

0.00%

171

2014

95.13%

4.54%

0.33%

216

2017

93.97%

4.58%

1.45%

190

 

66-75
Year Internet No internet Unknown Total

2011

65.83%

34.17%

0.00%

48

2014

84.39%

13.96%

1.65%

68

2017

87.82%

12.18%

0.00%

90

 

>75
Year Internet No internet Unknown Total

2011

57.14%

42.86%

0.00%

25

2014

62.33%

37.67%

0.00%

48

2017

51.71%

47.84%

0.46%

77

 

Total
Year Internet No internet Unknown Total

2011

88.87%

11.13%

0.00%

390

2014

90.02%

7.80%

1.00%

479

2017

84.63%

14.68%

0.69%

450

Did you find what you are looking for?

Your feedback will help us improve this website.

Thanks, do you want to tell us more?

Do not enter personal information. All fields are optional.