Discussion Paper No. 194 of 2017 on Digital Dividends: Evidence From Internet Usage in Kenya

Abstract

This paper examines the digital dividends arising from work-related Internet usage in Kenya by first evaluating the factors influencing Internet usage and, second, the effect of Internet usage on earnings from wage employment in Kenya. Robustness checks are carried out by conducting specification tests on home Internet use and specific Internet tasks. Using cross section data obtained from the 2010 National Information and Communications Technology Survey, the study used a sample of 3,783 working individuals between the ages of 15-64 years. The study used the Heckman Selection model to control for selectivity bias inherent in the Internet usage decision. From the empirical analysis, the study found that individuals residing in urban areas, those with many years of education, skilled labour and individuals working in the finance and insurance industry were likely to use the Internet. Regarding the effect of Internet usage on wages, the study found that individuals who used the Internet both at work and at home enjoyed higher wage dividends compared to non-users. Individuals enjoying these dividends were highly educated, in skilled labour occupations and residing in urban areas. Using the Internet for Internet banking activities was more rewarding, with higher wages than any other Internet usage task. The key policy recommendations are that the government could increase investments in Internet infrastructure to promote access by all people, particularly in the rural areas for uniform dividends. In addition to access, training of individuals on productive use of the Internet could be enhanced.