Let them eat internet!

Households making $25,000-$35,000 a year spend ninety-two more minutes a week online than households making $100,000 or more a year in income, and differences vary monotonically over intermediate income levels.

That is from a new NBER paper by Boik, Greenstein, and Prince.  Do note that the authors adjust for age and other demographic variables.

The upshot is that the real “undervalued” services from the internet come from its risk-sharing properties, not from the supposed lack of pricing of internet services.  If something bad happens to you, well…there is always the internet to fall back upon, at least provided you still can afford the connection.  This also means that business cycles are not quite as painful as before, but also that labor markets will be slower to adjust.

Some also may find in this fact an optimistic statement that “real life” (ha ha) has more to offer than the internet, with the caveat that real life is expensive.

The data in this very interesting paper also indicate that Chat has largely collapsed since 2008 as a way of spending time on the internet, internet time devoted to news sites has fallen from 10% to 5%, and social media and video are on the rise.

Here is my previous post “Let them eat ideology!”


How much of this is explained by the confound of younger households earning less and spending more time using the internet?

Right. And there's not really that much of a difference to explain, is there? After all, we're talking about a usage difference of less than 15 minutes a day.

"Do note that the authors adjust for age"

Isn't online time (including streaming and social media) today equivalent to tv time 20 years ago, or radio time 80 years ago? This also doesn't take into account high income earners who spend much of their work day online (although for reasons other than streaming and social media). The internet is a tool to some while entertainment to others. And so is a telephone. If the internet is undervalued, then so are the tv and telephone.

'If something bad happens to you, well…there is always the internet to fall back upon, at least provided you still can afford the connection. '

And if one pundit can be trusted, the U.S. will be creating Internet serviced favelas, places where the inhabitants do not even need to notice whether business cycles are not quite as painful as before.

They're called public libraries.

Good reason to consider the internet infrastructure, and to build a better one. This includes more seductive educational resources.

Physics lectures delivered by porn stars.

Better than the other way around

My PBS station had an immersive French show which followed the daily life of a French girl who looked good in a sweater.

A little more subtle than the YouTube "Hot for Words."

I don't know about you guys, but...I see interiority.

spending, infrastructure, coercion, education. i wonder if those suggestions are internet-specific.

Bad things were said in the break room. Bad things happened in seedy bars after 11 pm. But they were contained. They weren't accessible to teenagers with their cereal in the morning.


Free publicity

That callousness too was possibly seen by a few teens, as they process how adults think about things.

Good job. Make America Great Again.

Internet access is a fixed cost, so you do not have to spend for the marginal minute of internet usage.

Alternative uses of time may require the expenditure of money, whereas there is no marginal cost for internet usage, which explains why a wealthier person may use the internet less relative to a poor person, given the wealthier person has more spending opportunities.

Or, perhaps, poor people are smarter and therefor use the internet more or a different browser. ""Intelligence Quotient (IQ) and Browser Usage" was a hoax study allegedly released by a Canadian company called AptiQuant Psychometric Consulting Co. on July 26, 2011, that claimed to have correlated the IQs of 100,000 internet users with which web browsers they used.[1][2] Its claims that users of Microsoft's Internet Explorer had lower IQs than users of other browsers was widely covered in the media, and its revelation as a hoax was widely cited as an example of the weaknesses of the media.[3] The speed with which the story was reported was also alleged by some to be indicative of anti-Microsoft bias.[4]"

Does this take into account time spent online at work?

...are you somehow questioning the impeccable research of "Boik, Greenstein, and Prince" ??

If they say that $25K-35K peoples "spend ninety-two more minutes a week online" than $100K+ peoples -- that's the absolute truth across the nation.
It is not 91 or 93 minutes, nor 12 minutes -- it is exactly 92 minutes!

These guys are pros and can precisely measure internet usage versus income-level across a couple hundred million peoples.

Gave me a good chuckle

I think the answer to that is "no", because the paper "[uses] click-stream data for thousands of U.S. households."

Rich neighborhood: safe streets to walk in, beautifully manicured lawns and flower gardens to look at, golf courses to enjoy, hiking trails to explore, farmers markets to buy stuff at, bike trails to ride on.

Poor neighborhood: broken sidewalks where you have a nice view of drug deals going down, random gunfire at night, occasional roaming gang of teenagers looking for trouble.

Yes, I've lived in both neighborhoods and in the rich one it is much easier to leave the house.

Which one has more Pokestops?

It always seemed to me that one way Sci-Fi movies like "The Matrix" and "Dark City" went wrong is the assumption that humans could be forced by Aliens or malevolent robots to accept unreality as reality (to be batteries? emotional porno?). In fact, there would be plenty of volunteers. That is the underlying fear/fantasy which drives those movies.

Joe Pantoliano's character is an example of preferring unreality that was dealt with in the movie.

What is so surprising? People making 100+ or more go to the Opera. I watch Maria Callas clips on youtube.

Households with less money = Households with fewer people working
Households with fewer people working = households with more people playing video games and otherwise goofing off online.

I don't consume less internet because I have money to spend on ski vacations. I consume less internet because I'm too busy working.

Interested in how they broke time down, but not $5 interested. I suspect that those making $100,000+ are at "work" more than those making $25,000-$35,000, but that a portion of this "work" time may be composed of internet usage. I could easily see the former group being at work 3-5 hours more per week than the latter, but spending an hour and a half of that extra "work" time on Facebook or commenting on marginalrevolution.com, such that total internet usage could actually be the same (if "work" is its own category not sub-divided into "real work" and "screwing around online in the office").

To go along with this, $100k is more likely to be white-collar work at a desk with a computer and internet access. $30k is more likely to be service or blue-collar work where you don't have access to the internet.

This is probably less true than it was 5 years ago - even blue collar folks now spend time on their phones during the day (obviously some workplaces are stricter than others). I guess it depends on what you define as internet usage

In my day we jerked off to cave paintings.

In my day we had to use our imagination. We did not have cave paintings.

In Soviet Russia cave paintings jerk off you!

People greatly underestimate the value of entertainment. Think about those medieval cathedrals and music!

Boredom drives some to drink and do drugs. It is one of the reason idle hands are the workshop of the devil. Yeah for the Internet and BTW you can Internet free at libraries and other places. I have seen a guy who sits in a Burger King near me not eating but on the Internet with his laptop for long periods of time.

BTW to the HBD folks: The most enjoyed music in the world was invented by black USAers. Rock & roll, blues and jazz. What a great contribution to the world. Of course they add great grace and amazing abilities to sports.

BTW what we call poverty in the west is not a lack of money.

the internet is becoming television. naturally it would subsume everything and at last sadly it is subsuming passive entertainment....of course politics already became television, and as you are now witnessing the worse is tv, so the worse is politics as a deliberative process.

I would guess this effect is:

- Less time available to high earners generally, and they're more free with their money when it comes to decompressing, opening more non-net options (so yes "let them eat internet").
- Stronger social inclination, self confidence and higher attractiveness among the higher earning group. (The internet's great, but let's be real here.)

92 minutes per week is just under 13 minutes per day. This may be statistically significant, but it doesn't strike me as a particularly large effect size.

And that's per household, meaning even less per person?

I wouldn't trust the authors' data source enough to make any broad conclusions – including their own conclusions or Cowen's interpretations.

The authors describe their source as ComScore "click-stream data" from "primary home computers" (no more than one computer per household). They can't distinguish between multiple users of the same computer.

That leaves out mobile internet use and smartphone apps – large and growing factors by 2013. It leaves out smart TVs and game consoles.

It leaves out both personal and work laptops. It leaves out use on any additional home computers beyond the first. It can't see any "online time" (including for personal entertainment purposes) enjoyed at workplaces.

There's a good chance it only captures clicks from a particular instrumented web browser (if ComScore surveillance software is added to these 'primary home computers'), or those done with plain unencrypted HTTP (observable from an upstream appliance/relay). So using an alternate browser, or 'incognito'/'private' mode, or specialized apps (like ITunes) may all be missed.

Are smartphones, smart TVs, game consoles, laptops, multiple home computers, goofing-off surfing at work, or privacy-consciousness all potentially more prevalent among higher incomes? I'd say!

Higher-income individuals and households could be spending far more "time online" than those with lower-incomes – and this click-stream data would give no hint. In fact it might give the opposite indication, based on just the shrinking subset of activity captured by those single "primary home computers" with the ComScore browser installed.

Well said. I'll add tablets to your list of alternative ways to access the Internet that are more prevalent among high income households.

Comments for this post are closed