What is the consumer surplus from the internet?

by on March 8, 2013 at 11:41 am in Data Source, Economics, Web/Tech | Permalink

The Economist has an excellent survey article, here is one (not the only) estimate:

Yet another technique is to assign a value to the leisure time spent on the web. Erik Brynjolfsson and Joo Hee Oh of the Massachusetts Institute of Technology note that between 2002 and 2011, the amount of leisure time Americans spent on the internet rose from 3 to 5.8 hours per week. The authors conclude that in so far as consumers must have valued their time on the internet more than the alternatives, this increase must reflect a growing consumer surplus from the internet, which they value at $564 billion in 2011, or $2,600 per user. Had this growth in surplus been included in GDP, it would have raised economic growth since 2002 by 0.39 percentage points on average.

I would note one caution.  Consumer surplus per se does not make published gdp figures inaccurate for most purposes, since all goods and services yield consumer surplus to some extent.  One might argue, however, that the internet has higher than average consumer surplus, for purposes of thinking about human welfare.

Yancey Ward March 8, 2013 at 11:45 am

There are still only 24 hours in a day.

SR March 8, 2013 at 4:52 pm

Yeah, but the marginal utility of an hour of leisure is going up due to technological advances like the internet.

Margin March 10, 2013 at 4:32 pm


People get used to the quality of entertainment over time, they don’t feel much better per minute spent.

And the internet is often a place of hostile senseless pseudo-debates, where people hate each other’s guts afterward.

john personna March 8, 2013 at 11:55 am

The benefits of pervasive internet are massive. And people certainly do use those on-line hours to improve off-line activities. You can find better cooking techniques and new hiking trails. Those better pursuits may lower GDP and increase happiness.

Bill March 8, 2013 at 12:01 pm

+1 Substitution effects imply an increase in happiness, don’t they. Plus, if you increase your leisure time at work by being on the internet, you are increasing your happiness at the expense of your employers.

So stop reading these comments and get back to work.

john personna March 8, 2013 at 12:40 pm

For what it’s worth, I’m that other demographic … retired guy tells people what he thinks ;-) for a short term happiness blip.

Candide III March 8, 2013 at 8:18 pm

> Substitution effects imply an increase in happiness, don’t they.
Not necessarily. Crack got substituted for cocaine, heroin is frequently substituted for lighter opiates, vodka for beer and wine, antifreeze for vodka etc. Arguably, a lot of so-called consumer surplus from the internet is of this nature.

Doug March 8, 2013 at 3:08 pm

The problem is you can’t pay pensioners and national debt in the form of increased happiness.

In other words the CPI is not the same for everyone. If you’re a young, single, tech-savvy guy your standard of living has improved immensely relative to your nominal wages. Implying that your real wages have risen faster and that your effective inflation is low or negative. If you’re a mid-50s dad paying for a fair amount of healthcare, college tuition for a couple of kids, and a house in an expensive market like LA or New York suburbs, (add in that you’re generally befuddled by all things computers) your consumption basket has risen in cost much faster than the CPI.

In the past the deal was as technology advanced that lowered the cost of many items, in exchange the central bank got to offset that by printing money, ideally targeting 2-3% inflation. This kept the economy running smoothly and predictably, avoiding sticky wages, deflationary spirals, inflationary bursts etc. It also meant that in periods of faster technological advancement the central bank printed more money. Debt-holders received more diluted money, but the stuff they could buy with it was better. And vice versa in periods of slow technological advancement. The deal was if technology did well, we could print a lot of money and pay our debts easily without the bond markets throwing up. What Steve Jobs giveth Ben Bernanke taketh away.

Recently though the rate of technological advancement started fracturing across industries. You might call this “The Great Stagnation,” but the stagnation has certainly been a lot greater in healthcare than microchips. Even major items diverged, automobiles have had a lot lower rate of inflation than housing/construction. What I suspect is that traditionally the variability in inflation between different consumption baskets is at an all-time high. In 1960 pretty much everyone’s consumption basket improved and inflated at roughly equal weights. Whereas now, as with my previous example, different groups experience inflation at much different rates.

Overall this throws a wrench in macroeconomics, and I’m not sure what the right response is. Traditional macro says target 2% inflation. But what if half your population is experiencing -1% inflation and the other half is experiencing 5% inflation? You’re averaging 2% inflation, but no one is actually experiencing it. The former might have sticky wages and be stuck in an AD shortfall, but bringing them up to 2% would cause the other group to be at 8% inflation, which is miserably painful. Do you lower, raise or keep the same monetary policy?

john personna March 8, 2013 at 3:33 pm

The “pensioner” lifestyle was, for a time, quite something. It included vacation homes, multiple cars, and an unprecedented level of foreign travel. In contrast, in my youth “pensioner” meant someone in a worn sweater, in a small apartment. I don’t think we’ll fall quite that far again, but some adjustment seems inevitable. But there is nothing wrong with a simple, lower consumption, but civilized lifestyle. If cheap broadband makes that more civilized, why not?

Brian Donohue March 9, 2013 at 8:45 am

Great comment, Doug.

Rem acu tegisti!

SR March 8, 2013 at 4:55 pm

+1 too. Yup, and you know what else is a consequence of an increase in the marginal utility of a an hour of leisure _relative_ to the marginal utility of dollar of consumption? A drop in labor supply.

Dale March 8, 2013 at 12:09 pm

So, Apple is severely undervalued. Think of all the time people spend in their stores gawking at their products. Revealed preference says that they must value that time more gawking at Apple products than whatever else they were doing. Those social gains are attributable to Apple. If they could only figure out how to monetize them.

Wait, haven’t they monetized them already? Isn’t that part of why Ipads cost more than other tablets? For that matter, don’t subscriptions to internet services already reflect their value? Attributing further consumers surplus to the time spent browsing (or whatever) risks double-counting those benefits. It also risks making consumers surplus measures almost meaningless. Perhaps we should attribute additional value to the sequester on the basis of how many hours people spent reading about it or listening to talk radio?

Bill March 8, 2013 at 1:02 pm

First, The argument doesn’t follow regarding Apple valuation and Apple capturing surplus.

Apple is not the market, which means, Apple is a competitor in a market. If Apple has no market power, that if there are alternative competitors, they will compete, allowing you to capture the surplus. The rule would be the other way if Apple were a monopolist–it would be able to retain the surplus.

Second, you are correct that you should not be spending time listening to talk radio, unless it is NPR.

mavery March 8, 2013 at 1:06 pm

Are those social gains or has Apple just made it more difficult to optimize consumption? Browsing could easily be viewed as friction in the process of purchasing products. If so, Apple’s large market share has also resulted in lots of unproductive hours being used making purchasing decisions.

Ronald Calitri March 8, 2013 at 12:22 pm

The internet is not free! In fact, ISPs are already over-milking it, as all the comparisons with other countries indicate. Computers cost something, too, last I looked.
Don’t worry so much, nobody’s getting away with anything. The “providers” are hard at work keeping U.S. effective demand low.. Even in imaginary space.

Doug March 8, 2013 at 2:49 pm

My ISP charges about $50 a month for broadband Internet. I’d easily pay $50 a day for the Internet if forced to.

Phil March 8, 2013 at 12:49 pm

But doesn’t the marginal purchase have zero consumer surplus? Otherwise, we would borrow money and buy more things.

To say you’re overestimating the benefit of the internet because *all* goods have consumer surplus, you’d have to argue that all goods and services have consumer surplus relative to the days before the internet — that is, that everything is improving. Which is true.

Am I getting the logic wrong?

Doug March 8, 2013 at 2:48 pm

“But doesn’t the marginal purchase have zero consumer surplus?”

The marginal purchase can be an extremely thin edge for something with deep price elasticities. For example I’ve done about 10 Google searches today. I probably would have paid $50 for one of them $25 for the next 5, and $5 thereafter. The Google searches I didn’t do today (i.e. the marginal purchase) are worth $0 to me, otherwise I would have done them for free on Google already.

So the fact that my 11th Google search of the day has zero consumer surplus to me, doesn’t imply that I didn’t yield $195 of consumer surplus today.

Anon March 8, 2013 at 3:39 pm

Yeah, all that free porn is worth something.

Phil March 8, 2013 at 4:54 pm

Doug: agreed, the total consumer surplus is positive, even though the marginal is zero. My point is: if internet comes on the scene, and you buy some, wouldn’t it displace the purchase with the lowest consumer surplus, which would be near zero?

Silas Barta March 8, 2013 at 5:53 pm

You *might* argue that there’s been a consumer surplus … like everyone around has been doing for years, whilst you do everything you can to handwave it away.

John March 8, 2013 at 10:15 pm

I haven’t read the article but for the most part I think estimates of consumer surplus are largely useless academic exercises or purely of value for producers seeking to capture more of it. Personally I’d be much more interested in estimates of producer surplus.

I suspect both are significant but would like to see the strong competitive pressures on the latter rather than strong marketing efforts to capture the former.

John March 8, 2013 at 10:24 pm

I’m not sure the conclusion that the increase in time on the internet increasing from 3 to 5.8 hours/week indicates greater value of the time on the internet.

In 2002 watching video over the internet was a rare event and download speeds poor for just about everyone. Now, with Hulu and Netflixs we’re watching TV and rental movies that used to be consumed via VCR, DVD and Blueray. It’s not entirely clear that the cost of getting the rental movie was high as most were probably picked while shopping for other items. Moreover the shopping itself is a form of entertainment for some.

If the consumer was consuming the same product but delivery mechanism is the only change then it’s not clear the time spend on the internet is of any greater value that the time spent consuming the item a different way.

source March 21, 2013 at 6:55 pm

I have not checked in here for some time because I thought it was getting boring, but the last few posts are great
quality so I guess I will add you back to my daily bloglist.

You deserve it friend :)

Comments on this entry are closed.

Previous post:

Next post: