I think the idea is this:

Since politicians reward their favored special interests by means of exemptions to the income tax, if you change to a consumption tax you have removed a potent source of political favors. If the tax were fixed at a flat rate, then you wouldn't have a place to insert special tax favors. Even if they started putting in favors, they would have to be in the form of exemptions for certain types of consumption. It's harder and more politically dangerous to insert, say, a consumption tax break for buyers of multi-million-dollar yachts than it is to give that same demographic an income tax break. And folks only buy so many yachts, so you'd need a larger number of favors to get the same dollar value of special-interest goodies. Hence the politicians are more limited in their power to favor certain groups.

It doesn't completely eliminate such favors, but it might prune it back a little bit.

That is a generous interpretation of Wikileaks' intent. Others might see it as Assange on a power trip or political crusade. At the very least it's vigilanteism.

Yes it is. Wikileaks has no reasonable expectation for help from the government, given that the government believes the release of the documents would harm national security. If Wikileaks thinks its judgment is better than the government's, then it needs to back that belief up by doing the job right. If they can't even manage to filter out things that clearly should remain secret - e.g. informants' names - then they should leave it to the professionals. When people's lives are on the line, half-assed doesn't cut it.

I don't think most tea partiers advocate murdering politicians.

There is no perfectly secure approach if any communication into the power plant systems is required. (Getting data out you could do securely by radio transmission, but data in is problematic).

But I would argue it's easier to secure that laptop than to protect an online control system.

First, the laptop need not be connected 24/7 - you can connect it only for brief periods to the Internet (i.e. only while downloading the laptop's regular software/AV patches and whatever data you need to transfer to the power plant). That reduces the exposure considerably.

Second, in this scenario it is impossible for the control system to communicate directly with a potential attacker. Any attacker instructions would have to go through the laptop sneakernet. This is more difficult to make work for the attacker. It also places a latency penalty on attacks. That enables things like auditing the laptop before it's allowed to connect to the control system, thus giving the defenders a chance at discovering the attack before it can do any harm. You can't do that if the control system is on the Internet.

Third, even when on the Internet, a laptop is not easily identifiable as a piece of critical infrastructure to an attacker who has infiltrated your corporate network from the Internet. A power company LAN may have many, many PCs and laptops. It has far fewer routers, such as those used to control access to the control systems. Forcing the attacker to find the needle in the haystack (which may even be offline at the time) adds some level of security.

Being offline gives you a far better chance against attacks than being online. It's just inconvenient and more costly. Since the utilities face little market or governmental pressure to be secure, cheap and convenient wins over security.

Maybe. That is a risk, but I guess I see using the score improvement data as an improvement over using no data. I do think we need to measure other things too and be realistic about the limits of a given metric. I view using the present data as a start of a process of improving the evaluations, not the end.

"Teaching to the test": if what we want students to learn is not what we test, then clearly our testing scheme is bad. We need to fix the tests to be more than just regurgitating a few facts. I know that's easier said than done, but it seems like a big yet solvable problem to me. Not just for purposes of teacher evaluation, but for measuring student progress in general.

These are all valid issues. Some may be identifiable and corrected for via statistical analysis, some may not.

My feeling is that we don't need a perfect comparison function. We don't need the kind of precision to tell us whether Peyton Manning or Drew Brees are the better quarterback. Any team would be thrilled to have either of these guys on their team.

We just a reasonably objective measure that can sort the teachers into rough groups: the terrible, the average, and the stars. Most teachers and administrators probably already know who falls in what category. But the unions try so mightily to protect even the worst of their members that we need something objective that schools can use to overcome the inevitable accusations of bias or evildoing. That'll let us get the worst teachers out, reward the best teachers, and give the average ones a little extra data and insight into their students' progress.

Certainly no meaningful job, like teaching or engineering - can be boiled down to only one metric. But certain metrics are very important and should be a significant part of the evaluation.

For example, I'm a software engineer. My employer places a lot of weight on ability to perform development efforts according to a budget and schedule. These are not the whole picture - it doesn't measure quality, for example. And every development effort is unique, so setting the budget is an error-prone process. Often as a developer you need to deal with an inadequate budget or schedule. Sometimes you get a particularly tough assignment. You do the best you can. Managers realize these constraints are there, and you are not judged entirely on budget performance. But if you consistently fail to come close to budget, while your peers don't consistently have the same problem... that will be noticed.

Teaching seems like a similar set of constraints to me. Every student may be different, and standardized tests scores may not be the whole picture. But like a development budget, standardized tests do capture an important piece of information. It's not unreasonable for the customer - parents and taxpayers - to consider such things. Especially when taken over a few years' time where you can really start to see trends.

The value-added tests do also attempt to remove biases such as student selection, as the metric compares those particular students' scores against their scores from the previous year. So the metric measures just the kids in your class, and measures not where they started but how much they improved.

If the union were advocating that we measure additional metrics and publish those too, I'd be totally behind them. That way we could all debate how much we value the various elements of teaching, and see which teachers provide which advantages.

The problem I see is that rather than try to improve the objective measures available, they're trying to sink the use of such measures. There will never be a perfect metric of teaching effectiveness, just as there isn't one for programming performance. But the lack of a perfect solution shouldn't prevent us from seeking a solution at all. The status quo lack of any solution has not been serving teachers or students well.

The original LA Times article on the web did have a prominently placed solicitation for teachers to submit their comments on their score. Not sure what the plans for the dead tree edition were

It also seems to me that the teachers' side of the story was printed:

Whether you buy their arguments or not, the teachers' official point of view has been spelled out for the Times' readers.

I for one don't buy it. Certainly care needs to be taken with designing any evaluations of job effectiveness. The value-added approach tries to take such care. The union response was just the standard line that you cannot and should not evaluate them by standardized tests that would let you compare them against each other. And similarly disappointing rhetoric implying that only teachers can evaluate teacher effectiveness - as if mere mortals like parents or statisticians have no insight. You don't need to be a master chef to know whether the food was prepared by one, and you don't need to be a teacher to know whether a teachers's students are learning.

I don't think it means that, at least in general.

I do think it means you have to weigh the potential benefits against the potential costs of whatever regulation will come out of Congress or the FCC. Certain rules are great. Others, while well-intentioned, could hurt more than they'd help.

Part of that analysis is analyzing the proposed policy, and part of that is analyzing how the regulator will choose to interpret and implement the policy.

1. Policy:

Theft is clearly a societal harm. Defining theft is straightforward, so a law against theft doesn't also outlaw legitimate behavior. Thus larceny laws are a win for society.

Defining "net neutrality" is not that simple. The principle is that we don't want ISPs to interfere with their customers' choices of content or content providers. We don't want ISPs deciding for customers which types of traffic are "valuable" and which are mere nuisances (as Comcast did with BitTorrent).

But how do you craft a legal rule that does this but also enables ISPs to offer legitimate Internet services reliably (say, VOIP, videoconferencing, or other latency-sensitive apps)? Such services may require prioritizing certain types of traffic to function well. My 500 MB Windows Update download can tolerate all sorts of latency/jitter/etc. Many real-time applications cannot. So is the ISP being non-neutral if it enables VOIP etc. to operate smoothly on its network?

Other ISPs may not want VOIP on their network (since they also sell more expensive voice services). If they treat all packets identically are they then discriminating against VOIP?

This is a very tricky and highly technical issue. Many people are not sure whether it is possible to craft the rules to have it both ways. So the costs and unintended consequences of such a regulation may be significant.

2. The regulator:

Even if we somehow could craft an objectively adequate/beneficial neutrality rule, is that regulation likely to be what comes out of the regulator?

Given the technicality of the issue, I doubt Congress is capable. They simply lack sufficient technical knowledge.

The FCC is more technically inclined. But it has a long history of retarding technical innovation and creating barriers to entry into the market. Since past behavior is the best predictor of future behavior, it is reasonable to ask whether a "cure" devised by the FCC might be worse than the disease it purports to address.

My point is not to say we shouldn't consider neutrality regulations. The point is that all too many people on Slashdot are blindly assuming that passing anything labeled "net neutrality" will solve or improve the problems. This is not by any means certain, for the reasons given above. We should go into this process with our eyes open.

I agree with you about the spin Google's putting on their decision.

But it seems to me that Google has seen the political forces arrayed against neutrality, have concluded they can't get everything they want given the current balance of power, and are thus proposing a compromise.

It may be a genuine offer of compromise, under the theory that half a loaf is better than none.

Or it may be a scheme to divide and conquer the telcos by differentiating between wired and wireless ISPs. It removes significant wireless players from opposing wired neutrality now. And if neutrality gets imposed on wired ISPs, and a couple years later everyone can see no catastrophe has occurred, that will make it far easier to then eliminate the "loophole" for wireless providers.

But either way, this doesn't seem "evil" so much as Google recognizing the reality that they have been unable to persuade enough legislators that net neutrality regulations are a good thing.

And that's all assuming that the net neutrality regulations will actually ensure neutrality. Given the history of the FCC - protecting incumbents from innovation or competition - I wouldn't bet on it. The guys with the expensive lobbyists tend to win even when they "lose".

Show me the private delivery company that perpetually gets taxpayer funds to cover its losses ($3.5 Billion just for last quarter), and I'll show you one that delivers stuff anywhere for "44 cents".

Clearly their costs are far more than 44 cents a letter.

Don't get me wrong. The Post Office has some unique constraints, particularly its coverage of low-volume rural areas that don't get coverage from some of the private companies. So it's difficult to make a real apples-to-apples comparison. I'm just saying that 44 cent stamp is only part of the story.

Certainly the ISPs have limited accountability at the moment, whether through market competition or through the law. So the question is whether, by enabling the FCC to produce additional regulations covering this business, the ISPs' accountability to their customers will be increased or decreased.

My feeling on this is that their accountability will not increase. The telcos have a long history of skillful lobbying, or to view it another way, the FCC has a terrible record. The FCC protects incumbents from competition and innovation, rather than protecting customers from the incumbents. The EFF and Reason have some good documentation of this:
The FCC and Regulatory Capture
The Central Committee is in Session: The Trouble with the FCC

News happens within a context. Context is inevitably a product of the reporter's bias because it requires analysis and not mere recitation of fact. And attempts at summarizing complex issues often introduce bias because they require the reporter to identify which are the relevant pieces of information.

Consider those notorious pay-day loans to low-income people. These are usually loans that last a few weeks. The borrower is charged a fee of say $15 per $100 borrowed. Typical loans are only a few hundred dollars. If you express this fee as an annual interest rate (even though the loan doesn't last that long) it's a triple-digit rate.

This outraged advocates for the poor. So the government passed a regulation limiting such "interest" (really interest and processing fee) to "protect" consumers. The result is that instead of $15.00 per $100.00, the lender now gets $1.50 per $100.00, which isn't even enough to cover the cost of the paperwork on most loans, and the vast majority of these lenders are effectively forced to shut down and make no loans at all to consumers who need a temporary loan.

So what's the story? "Government passes consumer-protection bill" or "Government forces closure of businesses serving the poor"? Did the government help the poor or hinder them?

Sure you can present a nuanced view if you have enough time, but you A) still need a headline, and B) are often forced in the news business to make the presentation of the news too short to adequately convey that much nuance. The result is you get the best summary the reporter can come up with, which ends up biased by their viewpoint.

Then consider how that happens when the regulation isn't something small like just the payday-loan regulations, but a 2000-page behemoth like the health-care or financial bills.

You could note that the financial bill imposes some limitations on banks' risk-taking with their own funds. Or you could observe that stringent limits on that risk-taking were gutted by lobbyists and pressure from legislators from New York and Massachussetts.

You could note that the bill creates a "consumer protection" agency. Or you could note that it imposes regulations on all sorts of financial products unrelated to the crisis while doing little about the derivatives central to the crisis.

You could note that the bill imposes liability on the ratings agencies for making bad ratings, and that those agencies had contributed to the crisis by rating high-risk subprime securities as being low-risk "triple-A". Or you could note that the agencies are now so fearful that they have banned bond-issuers from using their ratings, and that since bond-issuers by law are required to include ratings in their documentation, that issuance of new bonds for consumer-type loans is way down this week because issuers are in a catch-22.

These are all true. Which of these facts are presented in the story can dramatically change the picture you get of the bill, and whether it is overall a good thing or not. Most of the media are running very short clips on any given story, so the distillation process ends up introducing a lot of bias. They may not state an opinion explicitly, but it's inherent to the selection of which facts to present.

Which is all a very long way of saying, you can't eliminate the opinion by just getting rid of brazen statements of opinion. You'll still have a liberal-leaning and a conservative-leaning media because their whole understanding of reality and what just happened is filtered through their ideological lens. And the kicker is they both think they're being totally fair and even-handed.

It seems like such a fundamental human problem that I don't see a solution other than getting news from multiple sources, with different biases. In a way, that's the silver lining to today's partisan media landscape. At least you know the bias of the organization presenting the news to you.

I would dispute that the regulators are accountable to the people. FCC bureaucrats have no public accountability. Elected officials, thanks to gerrymandering, have very little either (thus we have Congressional approval rates around 20% and re-election rates around 80%).

The way I look at it is, who do you believe will abuse censorship powers least? A group of politicians devoted to an ideology, or a corporation devoted to selling internet access for profit? Personally I see the politicians as being the ones with a greater incentive for abuse. The company just wants to take my money. Politicians want to take my money, AND tell me how I can behave and what information I should be allowed to see and hear and speak.