Andre Boik, Shane Greenstein, and Jeffrey Prince write (the link goes to an ungated but outdated version),
We find that higher income households spend less total time online per week. Our results suggest that a household making $25-35K a year spends 92 more minutes a week online than a household making $100K or more a year in income, and differences vary monotonically over intermediate income levels. Relatedly, we also find that the level of time on the home device only mildly responds to the menu of available web sites and other devices – it slightly declines between 2008 and 2013 – despite large increases in online activity via smartphones and tablets over this time. At the same time, the monotonic negative relationship between income and total time remains stable, exhibiting the same slope of sensitivity to income.
Think of allocating your time among three activities: work, online leisure, and off-line leisure (plus housework). Are we seeing some households choosing to work less and instead consume more online leisure (thus earning less income), or are we seeing households who earn less per hour worked finding offline leisure activities too expensive (Tyler Cowen seems to think it’s the latter).
How do we integrate non-market activity (home improvement, chores, child-care, elder-care, self-employment for self-consumption – i.e. anti-specialization) into this picture?
The prices of some traditional offline leisure activities does seem to have risen faster that inflation over the last 30 years. Sports tickets, movies, and amusement parks are some of my impressions of this.
Meanwhile my impression is also that the quality of the experience of these offline experiences has simultaneously declined, as they have become more crowded with worse crowds. Just like the average quality of flying has decreased both with additional security hassle, but also because airlines are more efficient these days at price discrimination and filling every seat, which makes for less enjoyable and comfortable flights.
Needless to say it is very hard to capture these quality shifts when trying to adjust for nominal inflation and estimate the change in real prices.
“online” is like “reading” – it could be hard work, self education, or leisure. But with online it’s quite straight forward for it to be two or three of those things at the same time.
(Would we try to measure work vs leisure by whether electricity or fossile fuels are in use?)
I think this data evaluation problem makes all conclusions drawn from such data very suspect.