This week for Ask an Economist I’m answering a question sent to me by Lawrence. He asked, “what [has] been the effect of getting rid of net neutrality?”
To answer this question, we have to wind back the clock to 2017. Then-chair of the Federal Communications Commission (FCC) Ajit Pai successfully led an effort to repeal a set of 2015 regulations on Internet Service Provider (ISP) companies originally put into place by the Obama Administration.
The simplest summary of net neutrality regulations is that they required ISPs to enable access to content on the internet at equal speeds and for equal costs. For example, your ISP charging you to get faster speeds on YouTube or blocking High Definition access on Netflix would be examples of violations of net neutrality.
The idea of paying your ISP extra to have access to certain websites is a scary one, but it appears the worst fears associated with the end of net neutrality were overstated. To some extent proponents of net neutrality are the victim of their own apocalyptic marketing.
The Rumors of the Internet’s Death Were Greatly Exaggerated
If you spent any time on the internet during the death of net neutrality, it was hard to miss it. On July 12, 2017 websites across the internet coordinated their messages for the “battle for the net.”
Websites including Amazon, Netflix, YouTube, and Reddit called their users to fight for “a free and open internet.”
On Twitter, #SaveTheInternet thrived, seemingly implying the internet itself was facing an existential threat.
After the decision by the FCC, CNN briefly declared it was “the end of the internet as we know it.”
Unfortunately for all kinds of doomsday prophets, extreme rhetoric always looks silly in hindsight when it fails to pan out.
Obviously we aren’t seeing ISPs charge users different amounts to use different websites in any systematic way. There’s no “pay your ISP to access Hulu” package yet. So already it’s clear some of the doom and gloom was over-hyped.
Fears of ISPs offering “fast lanes” to make users pay more for better service don’t seem to have materialized either. The only example of this sort of thing I could find was a Cox Communications trial of an “Elite Gamer” service. But this service was unlike the “fast lanes” hyped up by net neutrality proponents in that it never offered a less throttled experience and would have been permissible under the old net neutrality rules.
One of the biggest concerns about the repeal of these regulations was that it would lead ISPs to favor their own services. For example, AT&T owns Time Warner and HBO Max. In theory, AT&T could silently throttle speed to competing streaming platforms like YouTube and Netflix, thereby destroying competition.
So did this happen? Well it’s hard to say. ISPs don’t exactly release an annual report of internet traffic they throttle. But we’re not totally in the dark, either.
Researchers at Northeastern University developed a method to monitor throttling known as Wehe. The researchers tested data throttling before and after net neutrality, and the results are surprising.
If the researchers are right, ISPs do throttle services, but they were already doing so before the repeal of net neutrality rules. In other words, the repeal of net neutrality had little to no impact on throttling. You can check out the data yourself.
Net neutrality advocates may see this as a win because it provides evidence that ISPs engage in this process. But it seems to me this hurts the case for these regulations more than anything. Why?
The data, if correct, show that the internet net neutrality advocates fought for when they were campaigning to keep net neutrality was no different than the internet without those regulations. The internet that advocates were fighting to save already had throttling!
To me this is akin to someone who believes they’re drinking a Coke and they’re complaining about it being worse than Pepsi, only to be told they are actually drinking a Pepsi.
Another problem the data present is that most of the throttled data rates are fast enough to stream standard definition videos on YouTube, for example. Some rates are even fast enough to stream high definition.
Granted, different streaming services require different data rates, and some users have a strong preference for HD, but I personally have a hard time getting worried about being “throttled” into having my YouTube videos look like they did while I was growing up. Admittedly, I’ve never been obsessed with “graphics,” but 480p has always looked fine to me.
So, has the repeal of net neutrality been totally without downsides? Some groups claim there have been negative effects. For example, one report lists as downsides:
- ISPs are throttling data according to Northeastern University
- The $15 Gamer Fast Lane
- Real-time locations of consumers can be sold by cell phone companies
- Frontier Communications is charging mandatory equipment rental fees—including to customers who don’t rent equipment.
Points one and two have already been addressed. The Northeastern University Report shows the throttling existed before the repeal of net neutrality, and the gamer fast lane (which as far as I can tell, no longer exists) was also in compliance with net neutrality rules.
Point three may be concerning to some, but doesn’t clearly seem to follow from the net neutrality repeal. It’s not clear the government couldn’t address that issue separately.
And the Frontier Communications case was also already addressed legally without use of net neutrality laws.
Admittedly, there could be preferential throttling happening at increased rates that we don’t know about and perhaps other issues, but the verdict from what I can tell is that concerns about the repeal of net neutrality were enormously overblown.
Meanwhile, the EU, which does support net neutrality regulations, seems to have performed worse than the US during the pandemic. European regulators outright asked streaming services to throttle video speeds.
What is the Economic Reason for Throttling?
It appears the downsides from ending net neutrality regulations have been minimal, but is it true that ISPs have an incentive to throttle data? And is it bad if they do?
Proponents of net neutrality often argue against an ISP’s ability to do so because providing 5mbps speed on Netflix costs the same as providing 5mbps of speed on YouTube, for example.
But that isn’t exactly right. It’s true that technologically providing the same speed to different platforms is the same, but economically it is different.
To understand why, consider the market for groceries. Imagine Dan, Patrick, and Jon are the only three customers in the market for oranges. Patrick is willing to pay at most two dollars for an orange, Dan is willing to pay at most three dollars, and Jon is willing to pay one dollar. Let’s say the cost of producing an orange is 50 cents.
How much should the grocery store charge for oranges? Well, if it charges three dollars it will sell only one orange to Dan for three dollars in revenue and 50 cents in cost. Total profit is $2.50.
What if the store lowers the price to two dollars? Well, Dan still buys an orange and this time he pays two dollars. At the lower price, Patrick is willing to buy an orange for two dollars. The store generates four dollars in revenue and it costs them one dollar (50 cents per orange sold). This means the store has earned three dollars in profit, which is more than it got for selling the orange for a higher price.
Now what if the store lowers the price to one dollar? Jon, Dan, and Patrick all buy an orange for 1 dollar each which leads to three dollars in total revenue. Each of the three oranges cost 50 cents, for a total cost of $1.50. This time the store only makes $1.50 in profit. This is less profit than the two dollar option, so the store won’t be lowering the price to one dollar. Two dollars is the profit-maximizing price for the store. (These figures are summarized in the table below.)
|Price||Oranges Purchased (Dan)||Oranges Purchased (Patrick)||Oranges Purchased (Jon)||Total Oranges Purchased||Total Revenue||Total Cost||Profit|
Table 1: Grocery Store Profit With a Uniform Price for Oranges
So far in our example, we’ve assumed that stores only charge one price for oranges. That assumption isn’t a bad one. In many markets, there is a single stated price for all customers at some point in time. But it doesn’t have to be that way.
Imagine the grocery store could charge different prices to different customers for oranges. Now the store could charge Dan three dollars, Patrick two dollars, and Jon one dollar. In this case the store would make six dollars in revenue (3+2+1) and have a cost of $1.50 (0.50+0.50+0.50) for a total profit of $4.50. This is the best result for the store so far!
Economists refer to this practice of charging customers different prices based on their willingness to pay price discrimination. And price discrimination exists all over the place.
Senior discounts, first-class plane seats, algorithmic pricing, and grocery store coupons are all methods through which companies try to assess and charge customers based on their willingness to pay.
Is this a “bad” thing? Well, economics as a value-free field can’t answer that question, but it can give us some insight which helps us make our decision.
The first question we should ask is, “who most benefits from price discrimination?” There are two groups who directly benefit. Consider our grocery store example. Store owners are able to make more profit when they successfully price discriminate. This is one beneficiary. The other beneficiary is customers with low willingness to pay.
If the grocery store weren’t allowed to charge different prices to different customers, we saw that the profit-maximizing price was two dollars. Jon isn’t willing to pay two dollars, so he doesn’t buy an orange. In the world of price discrimination, though, the store is able to profitably sell Jon an orange for $1. Jon benefits from this, otherwise he wouldn’t have been willing to make the trade at all! The store will try to get Jon to pay as much as he’s willing to, but it will never charge him such a high price that he doesn’t consider himself better off for the purchase.
If there is a “loser” due to price discrimination here, it’s Dan. Without price discrimination, Dan would buy an orange for two dollars. With it, the store is able to charge him three dollars.
It’s important to highlight that Dan isn’t actually worse off for buying the orange here. He is still willing to pay three dollars for an orange, so by definition he values the orange at more than three dollars. The trade is still a win-win, but Dan would certainly prefer to have been charged two dollars, other things held constant.
So what can this teach us about data throttling? Throttling involves charging relatively higher rates for different services (either by increasing the price or decreasing the quality). In other words, throttling is a form of price discrimination. And, much like the oranges, providing speed to customers for different streaming or gaming services doesn’t mean the ISP pays more to provide it.
The oranges in our example above always cost the store 50 cents to produce. But charging different customers different prices allowed the store to make more profit, so there is an implicit cost to charging a single price (in the form of losing higher profits).
ISPs, like grocery stores, are able to earn more profits by charging different rates for different speeds on websites. This is no different than grocery stores benefiting from mailing out coupons to charge different prices for the same product, or airlines benefiting from selling first-class tickets to customers who want to pay for luxury.
Those less willing to pay for internet are more likely to be able to afford it when ISPs price discriminate. So much like Jon benefitted from price discrimination, some consumers would here as well.
As far as groups who would prefer no price discrimination go, it’s possible heavy internet users who spend a lot of time streaming and gaming online would pay more if ISPs could price discriminate.
But, at least for now, it seems like the benefit to ISPs of price discrimination is fairly low, as evidenced by the minimal impact of ending net neutrality.