Accessing the Internet: socio-economic and geographical determinants

Abstract

The rapid proliferation of the Internet has enabled people to access information and to communicate instantaneously with people all over the world, from almost anywhere, 24 hours a day. However, socioeconomic, ethnic, and geographic differences in rates of Internet access (known as the digital divide) persist despite increasing Internet penetration rates. Considering New Zealand's relatively high rate of Internet access, only a limited volume of local digital divide research exists. This study draws on the results of the 2001 and 2006 censuses to estimate the degree of influence that socioeconomic, ethnic and geographic factors have on access to the Internet in New Zealand. This study contributes to our geographical understanding of the digital age by showing empirically how apparent differentials in the geography of Internet access are actually due not to geographically based differences in the supply of Internet services but to the influence of socioeconomic and ethnic differences among households themselves. The argument is advanced in two strands. Theoretical expectations about Internet access are suggested on the basis of perceived benefits and costs to different types of households. These are then placed within a diffusion framework in order to derive expectations about different rates of take up by households with different levels of income and education. A major motivation for this approach has been to estimate differential penetration rates of Maori and Pacific Island households whose use of the Internet has been shown in previous studies to be considerably lower than that of Europeans. Multivariate regression and logistic regression models are applied to census data from 2001 and 2006 to establish the degree in which socioeconomic, ethnic and geographic characteristics influence a respondent's probability of Internet access. In theory, residents in urban areas are more likely to have access to the Internet than residents in more remote, rural areas due to increased cost of access and declining quality of service in remote areas. However the results of this study suggest that socioeconomic differences are a more significant determinant in estimating Internet access than geographic location; and while some settlement type variables are statistically significant, their influence in a multivariate regression model is relatively minor in comparison with the socio-economic and characteristics of households themselves. A significant result of this study is the influence that ethnicity has in estimating the probability of Internet access. Most of the overseas literature argues that ethnic differences in Internet access can be accounted for by their different socioeconomic characteristics as opposed to their ethnicity per se. However, this New Zealand study shows that, even when the major controls like income and education are in place, together with settlement characteristics. In a multivariate model, the influence of ethnicity on Internet access is still highly statistically significant. The lower access rates for Maori and Pacific Peoples respondents found in previous studies without controls are found to exist in a multivariate framework as well. The limitations of using census data to analyse Internet access is acknowledged, and an agenda for future research is established. With the overall Internet access rising towards 60 percent, future research might usefully shift from modelling Internet access to modelling Internet use; from studying who connects to the Internet to who uses specific Internet services. My principle argument would be that although access to the Internet has become more widespread, even among less advantaged populations the 'divide' has still not diminished. Rather differences in use are now compounding existing differences in access. Future research would assist in establishing exactly how different populations use the Internet and their relative educational returns to such use. Such an extension would return an even deeper understanding of the digital divide in New Zealand.