Posts by Thomas Lumley
Last ←Newer Page 1 2 3 4 5 Older→ First
-
One of the important benefits of proportional representation is precisely that splitting or combining of parties doesn't make much difference, so the whole concept of 'largest party' is about purely arbitrary boundaries.
Largest coalition matters, and obviously there's more negotiation overhead in putting together a coalition with more parties, but largest party just isn't a thing.
And, operationally, what would a "largest party" convention even mean? National should negotiate with Labour before they negotiate with NZ First, and with the Greens before ACT?
-
By personal and professional inclination I would have loved to believe in the Social Investment Unit, but not when it's run by a government that would make this sort of policy.
There's no way their expert advice can say meth contamination is a leading public-health problem. Or that drug dogs are a more effective use of public funds than Biosecurity Beagles.
I could see targeted use of searches in some special cases: there's some evidence from the US (PDF) that intrusive and unpleasant random alcohol testing of repeat drunk drivers is helpful. But if they had evidence that ad hoc removal of rights from gang members helped, they wouldn't have waited for the election.
-
Often I sit down to listen to music and then get distracted from really paying proper attention. But, judging from when I've been trying to learn songs, even the background music is getting in. Oh, and judging from the earworms. Right now, it's bits of Hamilton.
For at-home listening I do find that every time I upgrade my speaker system it turns out to be worth it -- but I haven't gone very far yet. Most of what I listen to doesn't need much bass, which seems to be where quality speakers make the most difference.
For travelling, I like my Shure noise-isolating earbuds a lot more than I expected to -- and it helps that plane flights give me time to listen without much distraction.
-
Your "Todays Q+A Report" link is being treated as a relative URL not an absolute one.
-
Money is safer on the Bitcoin blockchain than it is sitting on a bank’s computer system or when it is sent from one bank to another.
I'm happy to believe that, and it's nice for the banks, but it doesn't translate into risks for depositors: If Westpac is robbed or hacked I don't lose money, but there have been multiple examples of people losing Bitcoin because exchanges were robbed. As I'm sure Alex appreciates as a lawyer, it's the interfaces between blockchains and the rest of the world that are messy.
That's why I like Matt Levine's take on the DAO hack/fork as a complement to the techie ones. (there was a bug in smart-contract software; someone used it to steal lots of cryptocurrency; more than 51% of the community decided to edit the blockchain and change history so the hack hadn't happened]
-
Hard News: What the wastewater tells us…, in reply to Rich of Observationz,
Ritalin's not, but there is an amphetamine prescribed for ADHD.
-
While I agree it would be amazing if anyone paid list price for Sativex to get high, the Seattle cannabis shops (eg) do seem to advertise quite a few products with as much or more cannabidiol than THC, so there may be a chacun à son goût issue here.
It's possible that the high CBD products are for medical use, but there's a separate medical-marijuana network, so it would be a bit surprising if they all were.
-
Legal Beagle: New Zealand rockets up the…, in reply to David Hood,
Although I suspect it gives lie to Transparency International’s outline of the results as shocking because so many countries score below 50:
I'm going to partly disagree with David Hood and agree with Graeme here. The standardisation doesn't guarantee that a lot of countries will be below 50, but it certainly tilts the scale that way. Worse than that, it's possible that the proportion below 50 could increase even if all the changes that occurred were improvements (or the reverse).
The fact that improvements in one country will tend to lower the values (not just the ranking, which is inevitable) for other countries is more fundamentally wrong, though. Reducing corruption in other countries is good for New Zealand, not bad. This isn't rugby, where someone has to lose for us to win.
There can be good reasons for scaling an index differently each year. The motivation for 'grading on a curve' in the literal sense in large college classes is that each year's exams are completely different, so it's plausible the population is more stable than the difficulty of the test. If that's the case, scaling away year-to-year differences in mean and variance is sensible.Here, the idea is to have the same inputs to the index each year, and it's quite plausible that there are changes in the population -- that there are trends in corruption either around the world or in large groups of countries. In that situation it doesn't make a lot of sense to use a different scaling each year.
On the other hand, it is still true that given the actually-existing scaling, a country does have to be pretty bad to get a grade below 50. It's just that this is a contingent fact about the scaling, and might not be true in a different, better world.
-
Another interesting question is whether the Greens would ever be willing to compromise on zebrafish. These are used a lot in biological and biochemical research as a step between cell cultures and mice -- partly because they're transparent when young, but also because they're small, cheap, and about as cuddly as whitebait.
On the one hand, they're obviously animals. Vertebrates, even. On the other hand, there's much less social disapproval of being cruel to fish than to small furry animals. Think of catch-and-release fishing, for example.
-
Hard News: Greens drug policy 2: the…, in reply to Russell Brown,
could you use existing evidence for, say, cannabis or MDMA?
I’d say so. I mean, if MDMA were legalised I’m sure there’d be lots of interest in research on the more subtle or longer-term health effects, but we must know about as much about short-term safety risks as we would from a Phase II human study.