Posts written by Thomas Lumley (2585)

The Herald is reporting on the Government’s new road-cone hotline. Apparently there were 236 reports in the first four days, and the article usefully points out that this was about 100 for each of the first two days, so it dropped pretty fast after that.

A bit less usefully, there’s a graph (click to embiggen) of reports by miscellaneous administrative category . Auckland (region), as usual, is at the top.

It’s not completely clear what the right scaling is, but raw reports aren’t it. The traditional Kiwi favourite of per capita shows Wellington (city) way ahead of Auckland (region), with more than half as many reports from about an eighth as many people.  NZTA is there to represent state highways, and it doesn’t really have a population in the same sense as the other areas.

We might also scale by the amount of road to have cones on.  This map, from Figure NZ, isn’t quite what we want because it uses consistent geographical units — regions– but it does show how state highway and local roads compare. It looks like NZTA is coning above its weight per km of road

FigureNZ also has a graph by territorial authority,  a better match to the Herald’s graph, showing Auckland has well over twice the sealed road of any other authority (if you aren’t a Kiwi: this is ordered north to south).  Wellington has much less road, so it is attracting comparatively the most reports.

So, the data do show a “hotspot for cone concerns”, but it’s Wellington, not Auckland.  This could mean more cones, or streets with less spare room, or fewer alternate routes, or people who just whinge more. The data do not suffice to tell us.

• Ask Stuff’s BudgetBot anything you need to know about Budget 2025. Or, perhaps, don’t.  The budget is an occasion where there’s actually quite a lot of expert analysis both in mainstream media and social media, as well as all the unreliable vibes-based commentary anyone could want
• I’m not saying that AI is useless.  Mike Caulfield has some very impressive demonstrations of forcing Claude to help with fact checking.
• Various sources report that smartphones cause haemorrhoids.  This is based on unpublished research presented at a conference, of a sort that can’t show more than a correlation, and that just barely provides evidence of a correlation (this is the closest we have to details).  The problem is that the study has no ability to tell whether using a smartphone on the toilet increases the risk of hemorrhoids, or having hemorrhoids increases the chance that you’ll use a smartphone on the toilet, or that something else affects both things.  Also, it’s about time for this story to reappear, since we had it in 2023 and 2020  and 2018, though those times it was at least presented as expert opinion rather than research.
• RadioNZ have launched a political poll with Reid Research (hat tip to @danxduran on Bluesky).  Notably, the article describes not just the maximum margin of error, for proportions near 50%, but margins of error for smaller proportions such as 10% or 20% (not for 5%, unfortunately). These are uncertainty estimates for an idealised mathematical model of polling, and underestimate the true uncertainty a bit, but they are a big step forward.  I’ve written about the uncertainty for smaller probabilities on StatsChat before

There’s a new study of cancer and exercise that’s just been reported at a cancer conference in the USA and published in a major scientific journal, and which has made it to the media.  It’s good news; and actually real good news.

The study finds that exercise actually improves survival in colon cancer.  More precisely, it finds that providing an exercise coach improves survival compared to just providing the usual “exercise good; junk food bad” information.  Obviously this wasn’t a double-blind trial — you can’t make people exercise without them knowing — but it measured objective health outcomes. In particular, the medical profession can measure death very reliably.

There have been lots of papers in the past showing that exercise is correlated with better health in people with colon cancer. The correlation is robustly unsurprising: the less well you are, the less you are able to exercise, and there was no way to be confident anything more than this was going on.  This study was different, because people were randomly assigned to higher or lower pressure to exercise.  It’s pretty unusual to have a study that actually changes people’s level of exercise over a long period, and even more unusual to show that it actually improves their health. We don’t know if the effect translates to other cancers — previous studies have had hypotheses about mechanisms that are specific to colorectal cancer and others that aren’t.

Since this is StatsChat, I do want to compare what the research paper and the news said about the size of the effect. Here’s the graph from the research paper

At the planned 8-year follow-up point, the difference in survival was 7 percentage points. Basically the same was true at the planned 5-year point for survival cancer-free.  The overall survival difference narrowed a bit if you took the data out to ten years, and the cancer-free survival widened a bit.   You could also quote the average ratio of the death rates (or cancer recurrence rates) in the two groups, which is common in the statistical analysis of cancer but is a bit harder to translate into real-world impact (and which gives much bigger numbers 28% or 36% reduction)

The Guardian just reported the relative rates. The BBC reported both, very clearly.  Ars Technica reported both, but didn’t link the absolute and relative numbers as clearly as the BBC.

The Guardian also made a lot of the “better than a drug” comment by the chief medical officer of the American Society for Clinical Oncology,

“It’s the same magnitude of benefit of many drugs that get approved for this kind of magnitude of benefit – 28% decreased risk of occurrence, 37% decreased risk of death. Drugs get approved for less than that, and they’re expensive and they’re toxic.”

I think it’s worth noting that this is not saying exercise is better than chemotherapy. It’s saying exercise plus chemotherapy is better than chemotherapy along, and the margin is large enough that if it were new drug+chemo vs chemo alone you’d easily get approval.

From RNZ this week

The Wedding Planner director Susannah Reid said in 2023 $58,000 was a typical budget for a wedding across New Zealand but this year, the average cost was more like $87,600.

As we’ve seen in the past, this use of ‘typical’ by people in the wedding industry doesn’t have a lot to do with how the word is normally used.  In that post, in 2014, the typical value from the marital-industrial complex was $30,000.  In 2025 dollars that’s about $40,000, so apparently the real cost of “typical” weddings has more than doubled over the time StatsChat has been running.

Radio NZ also have this week: How New Zealand couples saved on their wedding (and what they splurged on). The most expensive wedding here is less than a quarter of the “typical” budget; the others are less than 10%.

If you think about it, estimating the actual cost of a typical wedding is quite hard.  Technically it’s not that difficult: the government keeps track of marriages, so you could do a survey of a sample of marriages and ask people, but the Births, Deaths, and Marriages people will only let you browse arbitrary marriages from the distant past. Contemporary marriages are public records, but public access to them is by name rather than just by year.   If you don’t have a genuine sample, you’ll tend to notice big weddings more than small ones.

Newsroom has an article about tertiary funding in the new budget.  There’s a graph showing how the government fee cap and inflation have competed over the past few years. The graph in the article (ignoring the impact of a fee-free year) is

I find this a bit hard to interpret — it’s not easy to see the cumulative effect of the two changes and work out whether fees have shot ahead or lagged behind.  I think this is clearer, showing the cumulative effect of inflation and the (maximum) fee increase; the highlight is the two Covid border-closure years.

We can see that fees had increased in real terms but had now fallen behind inflation. Based (obviously) on a projection for 2026, fees might be slightly ahead again.

Another option is to use the inflation series to deflate the fee cap and just show the real-terms fee changes. Again, fees were up in real terms, then down, and are right at the 2014 level. Again, this ignores the impact of a fee-free year, which will depend on when one started uni.

What I think this shows is that adding up yearly changes (especially multiplicative ones) in your head is hard, and it’s probably harder than the opposite task of estimating changes in slopes.  If you need both scales, you might be better off with a cumulative graph.   The big disadvantage of a cumulative graph is that the visual impression can be quite sensitive to when you start adding.

RadioNZ have How many years would you have to skip coffee to save enough to buy a house?

The story correctly says (a) a lot, even if house prices didn’t go up, and (b) that’s not really the question.

There’s one point they miss  which is important to a lot of the narrative on house prices and saving (perhaps because this is a personal finance article, not a housing prices article).

The reason for housing unaffordability isn’t that Kiwis aren’t spending enough on housing.  You could imagine a world where housing was readily available and affordable, but people in general, or some group of people, couldn’t buy houses because they were spending all their money on beer and holidays and not saving anything for a deposit. In a world like that, advising people to save might be useful (if they listened).

That’s not the problem in New Zealand.  Kiwis, collectively, are spending far too much on housing. If one person gave up coffee or avocado toast  to save faster it might help them a little bit.  If we collectively gave up coffee and avocado toast to save faster, housing prices would just increase faster to compensate.

Q: Did you see golf courses cause Parkinson’s Disease?

A: Maybe?

Q: Pesticides. In groundwater.  We should close down the Auckland golf courses!

A: Possibly, but for other reasons.

Q: Aren’t the pesticides bad?

A: There are some pesticides with reasonable links to Parkinson’s Disease, in particular paraquat (actually a herbicide) and organophospates. Most of the evidence in humans comes from farm workers, who get exposed to quite a lot of the stuff

Q: So these could explain the golf courses?

A: They could, but there are other differences between the people with and without Parkinson’s in the study, not just golf courses

Q: Some dude on Twitter said the cases were all near the Mayo Clinic and the controls were from all over

A: That sort of thing, but in fact the research paper includes an analysis with controls just from the same county as the cases, and the results aren’t all that different. It could still be that that the cases lived nearer the medical center; we can’t tell.

Q: So doctors cause Parkinson’s Disease?

A: More the reverse.  The researchers did make sure the cases lived in the right county at the time of diagnosis, but family history can matter, and the cases had seen a doctor more often in previous five years than the controls.  Also, doctors do cause diagnosis of Parkinson’s disease, which is what was measured.

Q: But it might be real?

A: Plausibly, yes.

Q: What about Auckland, then?

A: We get our water mostly from the Waitakeres and the Hunua ranges, which are not topographically suited to golf courses.  And most of the groundwater used in New Zealand is deep enough to be relatively safe.  And NZ is a bit fussier about what herbicides and pesticides we allow.

To be fair, 1News does say the answer is “no”, in compliance with Betteridge’s Law of Headlines

So, with a spread of wins across MyLotto and in-store purchases this year, do online tickets or in-store ones have a higher chance of winning?

Lotto’s line on this is also straightforward: “Your chances of winning Lotto stay the same regardless of where you purchased your ticket.”

At the end of the story is this line: The Lotto draw is live on TVNZ1 at 8pm tonight.

Watching the Lotto draw should really be the answer to a lot of these questions. With scratch-and-win or computer draws you might hope there was a mistake in the programming or the distribution of cards, but with the lotto draw? Look at the balls tumbling in the transparent washing-machine thing. Do they look as if they know where you bought your ticket? Do they look as if they care?

Another way to look at this: suppose there genuinely was some way to get an edge in  your Lotto chances and beat your neighbours. If you want secret advice like that, the web page of the country’s most popular news source is not going to be the place to find it.

Stuff has a headline: Supermarkets say groceries are actually cheaper here than in Australia and UK

The first point to make here is that the comparison is of a basket of products chosen by Foodstuffs NZ.  It’s … not impossible… that a basket chosen by an Australian or British retailer might give a different result.

The second issue is that the comparison was of prices before GST/VAT. Australia and the UK don’t impose sales tax on some of the items in the basket but NZ does. Since the difference between PakNSave and Aldi Australia was about 7%, the impact of tax on the comparison on the comparison is pretty big. Comparing the pre-tax prices makes NZ prices look lower. Is this appropriate? It depends on why you want to know.

If you wanted to know whether moving to Australia would make your groceries cheaper, then you’d want the actually existing prices that include sales tax.  If you wanted to know whether the evil NZ supermarket duopoly was exploiting helpless Kiwi families, you might want the pre-tax prices, since that’s where the hypothetical price-gouging would hypothetically be happening.

Foodstuffs are clearly focused on the evil supermarket duopoly question (taking the “Nay” side) so it does make sense for them to look at pre-tax prices.

From RNZ

New Zealand has rates of sexual violence against teenagers above the global average, ahead of even a badly afflicted Australia, according to new research.

and from the Herald

New Zealand has rates of sexual violence against teenagers above the global average, more than even badly afflicted Australia, according to new research.

The research paper is here, and I would guess there’s a press release out there somewhere. The Science Media Centre commentary is here.

The first thing to note is that the new research tells us absolutely nothing new about rates of sexual violence against teenagers in either New Zealand or Australia.  Like many studies producing global estimates, the research takes the numbers already available for various countries around the world and tries to combine them.  For countries without reliable data this can give us new information, but not for countries that already have reasonable-quality data.  So, the research tells us what we already know about the problem in New Zealand — it is bad.

The second thing to note is that while the number for New Zealand is higher than the number for Australia, the difference is much less than the study’s uncertainty estimate (‘margin of error’). The real conclusion is that NZ is about as bad as Australia on this front.  In fact, if you look at the figure for women currently aged 20-24 it’s actually lower in NZ (24.5%) than in Australia (25.4%).  The fact that Australia is just as bad shouldn’t reduce our concern about sexual violence in New Zealand; international comparisons don’t help for telling us whether we have a problem.

Doing a global study like this is useful to the extent that you can get estimates for countries without good data. The researchers tried to correct for various biases, such as narrow definitions of sexual violence in some of the component studies, different age limits, not including boys as victims, and whether the questions were asked face-to-face or not.  They didn’t have any data in some places, and so patched in estimates from nearby countries.  It’s not clear that this all worked.

If you look at the estimates from across the world there are some surprising numbers.  Wars are generally recognised as bad for all sorts of violence, but the number for New Zealand women currently aged 20-24 (25.4%) is substantially higher than the Democratic Republic of the Congo (17.6%) and South Sudan (20.9%) which have recently suffered actual civil wars, and Yemen (12.7%), which is currently in one.  It’s notable that this low estimate is identical for Yemen and Syria, Sudan, Saudi Arabia, Afghanistan, Qatar, Palestine, Libya, Iran, and Iraq — it’s an example of using data from one country when you don’t have data for another. These surprising results come with very wide uncertainty intervals: Yemen’s is 3.8% to 30.7%, and since most such intervals underestimate the true uncertainty the estimate basically amounts to  ¯\_(ツ)_/¯

There just isn’t much information about rates of sexual violence in many countries, so even if you cared about international comparisons, estimates for every country aren’t all that useful.  I’m a bit surprised the Lancet thought this one was informative enough to be worth publishing.  The best you can really say about the research paper is that it gives an opportunity to get headlines about this problem, and maybe potentially do something to reduce it.