Wednesday, 28 November 2007

I've been having an entertaining debate -> argument -> flamewar with Laurence Boyce on LDV on the notion of a national universal DNA database. Finally had a comment "moderated" (I think they mean "deleted") when I told Larry to "STFI noob" and er, something a bit more complicated to explain but was about as offensive.

Remember, kids:

Anyway, this thread is partly so Laurence can come here and comment in an environment which won't be "moderated" (promise), and partly so I can outline my concerns on the notion of a national universal DNA database (I'll refer to this as a "NUDNAD" from now on, because it sounds childish and rude):

My first objection is that the idea that any large bureaucracy can build, maintain and run a system as vast as a NUDNAD without data compromise would frankly be hilarious if it weren't so terrifying.

My second is that any government with access to a NUDNAD would be sorely tempted to scope creep it into The War Against Terror, the "war" on drugs, or just to sell the data in it to interested 3rd parties. The consequences of this could be devastating.

That's without even getting to the argument about whether it would be liberal or not to do it.

To be fair, there are a few possible benefits: it would become a lot easier to figure out all of the people who had been at the scene of a crime. That could be handy if the crime scene was visited by relatively few people, and if the identity of any of the people was in question.

However, the idea that a NUDNAD would facilitate personalized treatments on the NHS is a joke - sure, having your DNA sequenced might help, but there's no need to then put your DNA sequence onto a NUDNAD - the GP's computer system would be fine.

Anyway, what do you think?

Wednesday, 21 November 2007

V is for, uh... Truth

I commend to you the Ministry Of Truth's lengthy and brilliant vivisection of David Cameron's speech on rape. I keep forgetting about MoT, for some reason, despite the fact that it is one of the, if not the, best-written political blogs going.

Tuesday, 20 November 2007

If I were Vincent Cable, I'd say something like this

HMRC has today displayed exactly why entrusting this government with an identity card database would be a terrible idea. Can Mr. Darling categorically state that such a cock-up could never happen with the National Identity Register? He cannot.

And yet, he will blab that this latest incident is not his fault. Well, he's right. It's not. He can't have been in the Treasury long enough to set the policy and the senior staff that presided over this incredible blunder.

That honour lies with Gordon Brown. Another nail in the coffin for his alleged brilliance. Since his time in office begun, Mr. Brown and his government have displayed gross incompetence, appalling judgement and disgusting atavism across a whole spectrum of issues.

Simply, LABOUR ISN'T WORKING. Mr. Brown bottled an election because he was afraid to lose it. Now it's clear why. Mr. Speaker, we will put forward a motion of no confidence in this farcical administration at the earliest opportunity, and put an end to a decade of incompetence, waste and corruption.

Saturday, 3 November 2007

I don't often just post links, but this one has been nominated as "one of the best blog posts ever", and I have to agree: The D-Squared Digest One Minute MBA - Avoiding Projects Pursued By Morons 101

Thursday, 1 November 2007

Addressing Voucher Deficiencies

In a previous post, I outlined my two qualms with education vouchers:

1. The flight of the middle classes to better schools, due to reduced deadweight costs.
2. That poorer children require more expensive education, and yet are least likely to get it.

I would like to examine a different idea: reverse-auctioning children to the bidder who offered to teach them for the least amount of money.

You'd need to collect a bit of information about each child and hir parents - household income, geographical location, disabilities and so on, not more than about four variables. Then the children can be collected up into tranches of similar kids for consumption by schools.

The advantages of such a system would be:
1. The most difficult children to teach would get extra money to pay for their education.
2. Schools could specialize in teaching particular kinds of children without fear of collecting 100% difficult children - even if they did, the financial rewards would make it worthwhile.
3. That every child would get an "equal" education.

The disadvantages being:
1. Parents don't get a say in the matter.
2. Presuming there is an open market, new "schools" could easily buy tranches of children and then teach them really badly, skimming the money off for a profit.
3. It's impossible to capture the additional cash that some parents are willing to pay to get a better education for their child, leading to there potentially being less cash in the system as a whole. Simultaneously, parents would still be able to buy home tutoring, etc, for their children, partially defeating advantage(3).

Adding a performance bonus for successfully getting children to target educational milestones would help negate disadvantage(2).

Adding some sort of parental "veto" (which puts the child back into the pool for a second round of auctions) would help with both disavantage(1) and disadvantage(2). Since all schools should become equal over time, there should be little reason to veto a bid eventually. Repeated vetoing could lead to problems.

Anyway, I put the notion out there in the hope of some constructive comments.


ps. it occurs to me that a lot of the problems of vouchers would go away by awarding vouchers to a value inversely proportional to parental income - since poor kids are likely to do less well - however, I further suspect that schools would pick "the best of the worst", further clustering the worst pupils in the worst schools.

pps. a possible alternative to a bid-down auction could be a bid-up auction played in several stages - each child starts out with a ~£3000 fund. Schools make offers for children they think that they can teach for £3000. Any kids not taken up (all of them?) would have £1000 added to their fund. Schools bid again for the kids they want to teach at that price point, and so on up. Parents can either pick a school offer, or they can reject all offers, and accumulate half the £1000 for the next round. Parents can perhaps make 5 refusals. There are probably weaknesses in this approach, it's just something that occurred to me right now.