Three essays on high-speed internet: Home and mobile internet adoption, effect of internet's price and speed, and welfare evaluation

The Internet brings substantial benefits to American households. This dissertation studies the internet adoption of American households and divides its presentation into three essays as follows.

The first essay analyzes the effect of demographic and geographic characteristics of American households on the probability of home and mobile internet adoption. The data source of this essay is the Current Population Survey Computer and Internet Use Supplement July 2013. Using the binary logit model, the results indicate that 1) the internet adoption rates tend to be higher for households with better education, higher family income, younger adults, school-age children, and urban residence; and 2) the home and mobile internet models have different sizes of marginal effects for several household characteristics.

The second essay investigates the effect of the internet service’s monthly price and download speed and the household’s characteristics on the choice of home internet service. This essay merges the Current Population Survey Computer and Internet Use Supplement July 2013 with the Cost of Connectivity 2013 dataset. Using the mixed logit model, the results indicate that 1) households tend to purchase the internet package that offers faster download speed at a more affordable price; 2) the demand for internet services is price-elastic; and 3) households with better education, higher family income, and younger adults are more likely to purchase high-speed internet service than households with opposite qualities.

The third essay quantifies the amount of welfare improvement and predicts an increase in internet adoption rate from cheaper internet price and faster internet speed. The data sources of this essay are the same as those of the second essay. Using the log-sum difference method, the results demonstrate that households with better education, higher family income, and younger adults tend to have higher internet adoption rates and earn larger amounts of compensated variation than households with opposite characteristics.