In other words, income and education, even when the generational and gender gaps are accounted for, contribute to a very great extent to the differences in the levels of access to the later Net in Canada. These results are stable across the 1999 and 2000 survey results, with only a slight shift in the geographic variable. The urban-rural setting gains in importance, in 2000, although its direction remains relatively stable compared with 1999. Overall, perhaps the most important factor in predicting the likelihood of having access from home is household income. In both 1999 and 2000, this variable far outweighs education, gender and age, emerging as the most significant predictor of household access to the Internet.

Regression model with employment status

We wanted to see if employment status, given other demographic characteristics, is a significant contributor to whether or not Canadians have had recent access to the Internet or Internet access from home. Detailed results and tables are found in Appendix A.

Among unemployed Canadians, the trend is generally towards a negative correlation with Internet access, (less likely to have Internet access), although the wide variation in the data leads to inconclusive results. Nonetheless, in the 2000 survey the negative correlation between unemployed Canadians and Internet access in the past three months is significant.

When we maintain other demographic characteristics, including education, age, income, gender and location, part-time employment has no significant relationship to recent Internet access or Internet access from home.