Points of view contrary to my own

1. Here is Timothy Lee on the Comcast merger.

2. Will Obamacare help you live longer?  This result seems too rapid and too large to be attributable to improved access to health care, and out of line with other more general (non-policy) estimates.

Still, many people are touting this result.  In any case we are committed to providing you with a broad range of perspectives, including those contrary to our own, so do read these.  I am in the midst of travel and will not have a chance to read the health care study for a while.


No link to Steve Sailer? Wait, does that mean...he's not contrary O_O

Well, duh. Robin Hanson has at least openly praised Sailer. Tyler is *very* status conscious and he knows that hinting that he understands the way the world works is status elevating in that it relieves him of the humiliation of having to sound sincere while the party line. Speaking truth to power, however, is not only bad for your career, its also quite Gauche.

It isn't hard to read between the lines, though, and Tyler knows this.

"Speaking truth to power, however, is not only bad for your career, its also quite Gauche."

This really hit close to home. Before I left academia, I was under the (false) impression that academia was a place where ideas were rigorously tested and challenged, and we were obligated to speak truth to power. I tried to do that...and quickly found myself in a hostile environment. I'm glad I left.

Great thing about reading between the lines is that by definition there's nothing there, so you can just make up whatever you want. Priors etc.

#1 I'm sorry but this is just a poorly argued case for more statist policy. Timothey Lee clearyl hasn't been keeping up with Wired Magazine and the many new, just-over-the-horizon technologies that are about to immeniently disrupt the ISP market. In addition, monopolies provide a net benefit to consumers in the long run as monopoly profits give companies the financial resources to improve capacity. This is Econ 101 people.

The assumption that monopolies will invest their financial resources in quality improvements instead of rent-seeking or consumption by owners is every bit as realistic as the assumption that the state responds exactly to the median voter and is unswayed by either interest groups or bureaucratic self-interest. Which is essentially to say that both assumptions are terrible. There are legitimate problems with state regulation that should concern us - but arguing that monopolies are good for consumers makes no sense either from economic theory or from a historical perspective. As in most of economics we face a trade-off: is the harm from a monopoly worse than allowing state intervention? I don't know enough about internet technology to answer for this particular case, but pretending on either side that markets (even noncompetitive markets) are perfect or that state intervention is perfect is not a helpful way to frame the problem.

Econ 101 teaches that monopolies will earn monoply profits but surely these profits do not dissapear into thin air. They must therefore, almost by definition, represent money transferred from unproductive sources to productive sources. If one is concerned about monoplies rest assured that a technological, disruptive solution is usually on the way - monopoly profits incentivize creative destruction. Even Timothy Lee admits that the Bell monopoly was ultimately broken by technological innovation in the form of the internet. The messy 10 year court case to break up AT&T was completely unneccessary and represented a deadweight loss of resources. I think there is no trade off here, standing in the way of mergers is always bad.

I'm generally a big fan of applying Econ 101 straight to the real world, but it can be a decent starting point. In Econ 101 there are two basic problems with monopolies. First, they under produce the good relative to a competitive market because restricting quantity when you get to set the price is a profit-maximizing move. Second, they capture profits at the expense of consumer welfare. This might not be a problem if you believe - as you seem to - that those profits will be invested in expanding the business. There are two problems. First, in the absence of competition, the monopoly owners have no incentive not to spend that money either on personal consumption or on lobbying and establishing barriers to entry by other firms to insure continued profit. Second, the monopoly does not have an incentive to expand on quantity and quality because expansion is not the profit-maximizing move. In general, I find the assumption that monopoly profits will be used in a socially welfare maximizing way to be quite utopian. It is kind of like arguing from the liberal perspective that since tax revenue doesn't just disappear into thin air it must be the case that it's always being well used. In the absence of incentives to use the money for social welfare maximizing purposes, neither the government nor the monopolists will do so. (The government has at least partial incentives to increase social welfare, although I think it's correct to argue that these aren't strong enough).

Typo - as you might have guessed from context the first sentence should be "I'm generally not a big fan of applying Econ 101 straight to the real world...."

But this is not like taxes, Comcast is deriving income from providing a service in the market and should be rewarded. If it makes the smart decision to buy Time-Warner so it can temporarily charge higher prices than there' really no issue. This is literally transferring money from extremely unproductive people (people sitting around watcing Netflix or Youtube videos) to actual productive executives, lawyers, lobbiests, investors, etc.
You can't have the government regulating against clever market plays.

Note that you've switched from arguing that monopoly profits will enhance consumer welfare to arguing that monopolies deserve a reward for providing service to the market. The question of who deserves what could be debated forever, but it strikes me as odd to think a reward is deserved if the monopolist is maximizing profits at the expense of consumer welfare by under producing the good or by lobbying D.C. to get laws passed that create barriers to entry by other firms. (As to the tax analogy - the main point was about the realism of the assumptions. But it's worth noting that there's nothing implausible about the idea that government provides services in the form of public goods - this is the econ 101 way of looking at government. As I mentioned above, the assumption that government does that in strict accordance with consumer preferences doesn't hold - but neither does the assumption that monopolists act to increase consumer welfare, which has never even been a part of econ 101. The incentives just aren't there for the monopolists to do so.)

But even if the profits are not reinvested this is STILL no argument for government intervention to stop a fair merger deal. If anything monopoly profits will encourage creative destruction, so even if the company doesn't reinvest a penny it will still spur new technological development. On top of that the executives of Comcast worked hard to put together this deal, even if it comes at the expense of consumers it is unfair to deny glittering prizes to hardwork.

The fairness of the merger is context-dependent. Namely, it might seem "fair" for the company, while it's wholly unfair for consumers. It is impossible for the public to be prima facie *wrong* or *unfair* to challenge the merger's prudence when the parties affected by the merge do not include only those consenting to the transaction.

Further, 'creative destruction' isn't a reliable response. Creative destruction isn't an economic guarantee, it's a hope that while consumers are unduly burdened, someone will eventually fix the problem, knowing that there will be political and legal challenges along the way. I think it's amazing that, according to you, the entity that fixes the problem can be anyone except the government. The reason we have anti competition laws in the first place is to fight these merges so we don't have to rely on dubious counterfactuals to eventually fix the problem! Government *is* the market solution.

So, consider this counterfactual: instead of hoping that disruptive technologies one day come, the government busts the merger, and those folks that would have gone on to offer a new net service instead go on to build cheap, flying cars. If we're talking about alternative universes, which we are, I prefer the one I described to yours.

The only interesting part of this comment chain is that Nate K continues to joust with you as if you're not Stephen Colbert.



It's okay. I was done. I do feel somewhat silly now though.

Confession: I like feeling like I have a voice, even if the circumstances are a bit silly.

How magnanimous - not one, but two examples of contrary views.

The important point from Timothy Lee's article is not his point on the Comcast merger, but on how the Internet works. The Internet's model of Internet traffic exchange has delivered competition, that drove prices down. A change of that model to a model without competition is something to be worried about. Do know that for many years countries without a liberalized telecom market and former monopolist operators have tried to push this model through at the UN level and so far were unsuccessful.

But there are always disruptive technologies over the horizon and so ultimately it is quite foolish to be worried about monopolies. In the long run these sorts of mergers always benefit consumers and most of the concern is nothing more than anti-business mood affiliation. Productive members of society should not worry about small, temporaty price increases on their cable bills.

Try this argument: There's video and everything else. Video is what provoked this issue. Not (just) because Comcast wants more money, but because they are cable TV providers. They are losing business to cord cutters who just download the three shows they watch each week. They need to make up that revenue. So lobbying, lawsuits, etc.

But outside of video, the rest of the net is unaffected. In fact, because video is such a data hog, you would have to cut data maximums down below 4-5 gig a month to even start affecting non-video use. The other 99.9% of net sites will be fine, even as Comcast and others throw their weight around trying to cut off (or get more cash from) video streaming.

While it's true that a new app involving video would feel the effects of higher bandwidth prices, it's hard to imagine what that app would look like (telepresence?) And any other new use of the net would be unaffected.

The important point from Timothy Lee’s article is not his point on the Comcast merger, but on how the Internet works. The Internet’s model of Internet traffic exchange has delivered competition, that drove prices down

In most countries other than the US, the company that provides last mile services (ie. Comcast, Verizon etc.) is not permitted to be in the ISP business. Thus the consumer has to buy access and also pay an ISP.

I know of no such countries. Many countries, inspired by the European Union, have Local Loop Unbundling and Wholesale Broadband Access. The historical incumbent offers ISP services over its copper network, but also has to provide competitors access to these lines. I know of no country that has enforced structural separation in the telecommunications market. The UK has functional separation between BT and Openreach. Italy has discussed structural separation. The only structurally separated markets I am aware of are some railway networks and energy networks.

Following the inequality issue....everybody can afford a TV but not good teeth or education http://www.nytimes.com/interactive/2014/04/30/business/100000002855500.mobile.html

Your point is.....?

How can TVs have come down more than 100% in price? Are WE being paid to obtain them?

I was wondering about that too. It's probably as a percentage of current prices. E.g.. a TV that cost $1000 before and $500 now has dropped in price by ($1000-$500)/($500) = 100%. It's a bad way to measure things, but that's probably the explanation.

It's also flat out wrong to say they have come down 100%. You could say they used to be 100% more expensive. I suspect chicanery more than idiocy when this happens, because one must at least be aware of the option of using either 1000 or 500 as the denominator, and must be aware that they come back with very different percentages. It is quite implausible that someone who knows those two things cannot understand that the denominator is the reference point (old price, new price).

If you took the time you would see that it is a percentage of the average movement of prices through the time period.

Universities and colleges s are extreme price discriminators secondary school is paid for by taxpayers.

The health care study is well designed. Its value add is from looking at the population-level impact, beyond Medicaid, and the use of a natural control group in the over 65 population (who are generally on Medicare and did not realize any decrease in mortality). Also, I seem to remember there was a lot of fighting between researchers about whether the OR Medicaid study was even powered to detect meaningful but small changes in a number of health outcomes--this study doesn't have that problem.

And yet five year cancer survival rates in the UK's NHS are 50% compared with 70% in the US.

For how long, considering Obama's choice to head Medicare/Medicaid said he "loved" the NHS.

Or perhaps the Massachussets model is closer to Canada, which had 29 PET scan machines in the entire country compared with 2100 in the US.

A PET scan changes the diagnosis and treatment of cancers in half of the patients scanned.

The newly insured have access to PET scans and live in the country with the most scanners per capita. The expansion of insurance will only increase the survival rates, which by the way were much lower for the uninsured. USA USA USA.

Also, this says the cancer survivorship rate is 50% in England and Wales over 10 years, not five. Pretty good!


I would be plausible--perhaps almost inevitable--that RomneyCare would improve five year survival rates, simply by detecting cancers that won't kill you. But it also improved cancer death rates, which, in line with Tyler's notes, implies either an implausibly high efficacy of chemotherapy and radiation, or a problem with the data; benefits from early detection should not be showing up in any substantial way in Year Two.

It's not just about cancer. And I don't quite understand the logic of "almost inevitable–that RomneyCare would improve five year survival rates, simply by detecting cancers that won’t kill you". Detecting cancers _that won't kill you_ will inevitably improve death rates?

Your rebuttal only makes sense if you think medical technology has already peaked. If you don't think that, your rebuttal only strengthens Chip's point.

Quite the opposite. It doesn't matter if it has peaked or not. Chip's point is that socialized medicine--which he conflates with the expansion of coverage in the US--is bad for outcomes and that technology adoption is apparently a surrogate for that.

Comparing survival rates is generally not a good measure of the efficacy of healthcare regimes.


The Incidental Economist also has an interesting discussion comparing the power of the OR and MA studies, and I expect will continue to have useful posts on these two studies. With no disrespect to the authors here (I enjoy the eclectic nature of their interests), if you are interested in healthcare economics, TIE is a very useful resource.

Yep. Tyler's dismissal is pretty odd... the only reason being that "it's not what I would expect" certainly smacks of mood affiliation - slightly amusing / disappointing for those who remember his reaction to critics of the Oregon experiment, which as you point out had limitations not present in the MA study despite being similarly well designed. I'm sure he will update his marginal beliefs appropriately after reading it though.

That is one possibility. Another is that the outcomes that improved are not the ones researchers expect to improve with increased health care consumption, at least not at the observed rate.

A reasonable person would expect that same comment in either case.

A point to consider:

Patient A: Is clutching his chest, he is in the middle of a heart attack!

Patient B: Is a 30 something guy, feels fine. Starting to slack on exercise, his bad eating habits are lurking but only nitceable if you scrutinize his blood work.

Let's say everyone got insurance 24 hours ago. Chances are patient A is not going to get dramatically different care because of it. It's possible patient B will get dramtically different care, but the results of it will not be noticeable since he's a relatively healthy person If you decide to measure the outcome of insurance a single day later you will 'discover' health insurance has little or no detectable effect.

Now say you measure ten years later. The impact on Patient B's will start to become really noticeable IMO. So we need a sampling here not just of patients but of time. We need to see outcomes after extended periods of time. Unfortunately Romney care happened a finite amount of time ago so that puts an upper limit on the data available.

But other proxies could be used. For example, you could compare those who had uninterrupted coverage for the last 10, 20, 30 years versus those who didn't.

I would be plausible–perhaps almost inevitable–that RomneyCare would improve five year survival rates, simply by detecting cancers that won’t kill you.

This may be true, nonetheless in general detecting cancers earlier probably does improve outcomes. Even if you are also capturing people who simply lived knowing they had a cancer versus actually treating a cancer people didn't know they had in a productive manner.

> 1. Here is Timothy Lee on the Comcast merger.

This was actually an interesting and substantive article. But being Vox, it presents one side of the issue only.

And it presents the INCORRECT side of course

I agree -- I've read a few pieces by Timothy Lee on the peering issue, and I'm not impressed. Traffic ratios have always been an integral part of the determination on whether or not two carriers would interconnect on a settlement-free peering basis (just like scope of the networks is also a traditional determining factor). That isn't news and I would object to the characterization that some network services providers are choosing to arbitrarily impose new requirements on companies like Cogent and Level 3.

I also think it's troubling (but probably not surprising given the people advocating the regulations) how the entire last-mile net neutrality discussion is beginning to morph into a discussion about regulating the backbone. Mission creep...

Not sure I understand your argument. If I pay my ISP for X mbps of service shouldn't they provide what I paid for? Comcast is saying they don't feel like buying the backbone capacity to pay for the bandwidth they sold their customers. What am I missing here?

I'm sorry, but you are incorrect. Traffic ratios generally are not part of peering deliberations. Just look at the article Lee links to: The folly of peering ratios. You can also derive this from the member lists of European Internet Exchange Points and the participants in routeservers of these exchanges. In addition you can evaluate a large number of peering policies of ISPs that do not state a peering ratio, though quite a few do. PCH evaluated 144,000 peering agreements for an OECD study and found that 99,5% were done without a written contract and on a handshake basis, therefore without ratios. There is also no logical support for the idea as traffic going one way or the other doesn't cost anymore and the balance in traffic is irrelevant as equipment needs to be bought on a peak transfer capacity. So if your peak traffic is 9.5Gbit/s you need a 10Gbit/s interface (actually more), regardless of the traffic going the other way being 10:1 or 2:1 or 1:1.

In addition traffic ratios were in the past used by the likes of Level 3 to see who was big enough to peer with them, stating that the maximum difference was 3:1 or something similar. Now it is the other way round with the likes of Comcast stating that anyone sending 3:1 or similar to Comcast has to pay. Just shows that in a decade the market turned around from transit oriented to access oriented.

"The Folly of Peering Ratios" is a position document. It presents arguments for peering ratios and then it presents counter arguments against them. I don't read it as the proof of your position that you suggest it is. I will stand by my original assertion: traffic ratios *have* historically been an integral part of determining whether or not to peer on a settlement-free basis.

Regarding the figures you mention (144,000 peering agreements, and 99.5% being handshake agreements): I would guess those are collected from smaller ISPs that are interconnecting with each other at open peering IXs? That's the only way I could account for those types of numbers. In terms of Cogent-Comcast: are you suggesting that interconnection agreement is a hand-shake deal and that traffic ratios is not one of the commercial terms? I doubt that's the case based on the reporting.

Regarding whether or not there is a "logical support" for traffic ratios: that's an argument that is up for debate depending on what side of the traffic imbalance you sit. Clearly in the case of Comcast, they don't find it commercially advantageous to add ports on their interconnections with the likes of Cogent, when Cogent is massively out of balance (and Cogent has a history of these types of disputes too). Let the market decide how to cure that problem (as it has with Comcast-Netflix ...).

I've said elsewhere: Timothy Lee is the Rush Limbaugh of the Internets. He is about generating outrage. He knows what the average geek wants to hear, and then tells them how it's exactly right and that their geniuses for believing it. Oh, but watch out for those shadowy evil interests lurking in the corner!

good analogy

Timothy B Lee's name is unfortunate. I can't read him without repeatedly thinking he invented the World Wide Web, even though I know he didn't.

Ha! Thanks. All this time I had been assuming that this guy invented the web. I was surprised that such a luminary was writing these articles, but too stupid to realize he wasn't...

Compare all medical costs per patient in the States of California and Massachusetts. They are almost adjacent to one another on the national cost of living index (with pesky RI in between). MA spends more than three thousand dollars a year per person than California. Given its 6.6 million population, that's over $18 billion "extra." That's to save 320 lives, or $57 million a life. This does not account for improving pain levels, functionality, etc., but it's a figure that suggests Massachusetts citizens (Massachusettites? Massachusettians?) pay way too much of health insurance.

Comments for this post are closed