Archive

Archive for the ‘Statistrickery’ Category

Violence most horrid

October 28th, 2008 John B 4 comments

It only seems like a few months since the biggest “idiots failing to understand numbers” issue was that of crime… oh, for those heady pre-recession days when people had nothing more to worry about than lamenting that crime was a major problem despite it generally not being one.

The latest row, which has unfortunately given ammunition to the ‘all crime stats are lies because my uncle’s neighbour’s nephew’s dad’s dog got STABBED! with a KNIFE!!!’ brigade, is about police forces wrongly classing minor-ish violent crime committed with the intent of inflicting majori-ish violent crime as ‘other violent crime’ (in line with how the courts treat these offences) instead of ‘most serious violent crime’ (in line with how the statistical guidelines say you should treat these offences).

Mark Eaton has an excellent piece on this, using the old stats, the new stats and BCS data to make the point that the reclassification changed nothing significant (not least, because only criminal justice wonks look at these subcategories anyway, with reporters and politicians alike dealing solely in ‘reported crime’, ‘reported violent crime’, and specific types of offence). Hooray for the BBC for hiring him and saving me a job…

Categories: Bit of politics, Statistrickery Tags:

Furniture, nudity thereof

October 26th, 2008 John B 5 comments

If anyone uses the phrase ‘the cupboard is bare’ when referring to the UK’s current economic position, this is an excellent indicator that they are entirely clueless about it. For one, it’s an embarrassingly trite and twee metaphor, unlikely to be used by anyone literate; for two, it paints an entirely false picture of the government’s financial situation (also, note that the government precisely and exactly did pay down the national debt as a proportion of GDP, which is the only figure that matters, during the boom times, which makes criticising them for not doing so particularly weird).

As Chris highlights (while also, correctly, pointing out that the current ’sterling crash’ isn’t one at all – if it were, the pound would have fallen significantly against the euro, which it hasn’t. Rather, it’s a dollar/yen rally), people are still falling over themselves to lend the government money at very low interest rates. That’s an indicator that outside of cutesy-talking-point land, serious people accept there’s plenty of, err, cupboard-room.

While I’m here, a couple of points on the Centre for Policy Studies report that purports to show the UK has a ginormously terrifying public debt. For a start, it’s written by a Tory MP – aren’t think-tanks producing this kind of report supposed to maintain some vague pretence of not being entirely motivated by partisan hackery?

Content-wise, it’s the same report the CPS churn out every year, with the figures slightly updated. And as always, it’s spun ridiculously: the angle is approximately “when you include the PFI Enron accounting, Network Rail’s nationalised in all but name-ness, bank bits, various other dodges and public sector pensions, the national debt is enormous”. In fact, the only non-trivial sums it identifies are PFI payments – which it exaggerates by a factor of more than three by failing to follow anything even vaguely resembling accounting standards, as I’ve already pointed out here – and public sector pensions, which are an order of magnitude larger than any of the other factor, and are the only way authors of this kind of paper can get from “the national debt is 42% of GDP instead of 39%, nobody cares” to “oh my god, the national debt is 150% of GDP and we’re all ruined due to Evil Labour”.

Quite how the hell public sector pensions should be accounted for is a tough question, and not one which has been satisfactorily resolved anywhere by anybody. However, suggesting that the UK is particularly screwed because of Labour’s incompetence and dodging, when actually the problem has existed forever and in every developed economy, is grossly dishonest. It also doesn’t represent debt in the sense of ‘people who have pieces of paper saying you’ll pay them and who’ll sue you if you don’t’ – it’s just a promise from politicians to be nice to old people, and we all know about the iron-like unbreakability of politicians’ promises…

[*] Yes, Network Rail’s GBP20bn debt should be included in the headline figures, as it’s government-guaranteed and not secured against tradeable assets. So should the real PFI number of c.GBP30bn; together, these add an extra 4-5% of GDP to the official national debt figure. I’m happy to confirm for the benefit of readers who question my political neutrality that these should be classed in the national debt proper and that Labour are slippery sods for not doing so (although on the other hand, they were the first ever UK government to move to GAAP for public sector accounting and are one of the first globally to adopt IFRS).

The banks shouldn’t be: it doesn’t make sense to view debt backed by tradeable financial assets as part of The National Debt, since it doesn’t represent money that’ll have to be paid back out of future taxation. At worst, we’re on the hook for the difference between the value of the banks’ mortgage books now and the long-term value of the relevant houses, cushioned by homeowners’ wiped-out equity. Even if we have a two-year depression and house prices fall 40% from their peak, the loss potential isn’t high.

In which your host is proved right

October 21st, 2008 John B 10 comments

Every prediction I made in this piece from 2005 on 24-hour drinking has proved to be correct: on-trade alcohol consumption has fallen, levels of alcohol-related crime haven’t changed; pubs haven’t made any extra money; but puritan idiots have continued to rail against the rule change anyway.

The most offensively stupid puritan argument is that ‘24-hour drinking hasn’t cut violent crime, so it was a failure’. No – the point is, it means that law-abiding people can go out for a drink without having to obey insane rules created to stop soldiers in the trenches getting jealous of civvies back home during WWI. That is a good thing in its own right. If drink-related violence had risen, we’d need to weigh the good against the bad. Since it hasn’t, we can say that the licensing law changes are an unequivocally good thing, and crack open some booze to celebrate. Hurrah!

(it’s also worth noting that on this issue, the Tories are lying crooks who should be run out of town on a rail. Quelle surprise…)

Anything the global financial system can do, local government can do worse

October 8th, 2008 John B 3 comments

Individuals who lost more than £50,000 in the Landsbanki collapse certainly let greed get in the way of good sense, and certainly don’t deserve the generous bail-out terms that the government has given them. However, that pales into insignificance compared to the 20+ local councils who’ve lost tens of millions between them in Landsbanki deposits. And who won’t get a penny back, as compensation schemes for bankrupt banks only protect retail investors.

These organisations actually have people employed with financial qualifications in working out what to do with their money. And it’s not like they haven’t been burned before by the collapse of a dodgy bank that just happened to be the highest interest payer (if it is in fact possible to work in local government finance without being told about the BCCI collapse and its knock-on effect for councils, then there’s a systemic problem in that everyone in the entire industry is completely inept).

It’s unfortunate that local taxpayers can’t recover the missing assets from the idiots in question, and the councillors who’ve singularly failed to oversee them (and who, I’m willing to stake near-Landsbanki-style amounts of money, will be more or less equally drawn from the ranks of the major parties).

Update: The Daily Mash calls it right: “oh fuck, we meant ‘Luxembourg’”…

Sir Ben of Goldacre

August 30th, 2008 John B 4 comments

Buy this book. If you understand why you need to buy this book, then buy this book. If you don’t understand why you need to buy this book then – for the love of all that’s worth a damn – buy this book.

Just don’t listen to the author talk, because he’s got an unfortunately whiny voice – one of those chaps who makes those of us who’re ‘prettiest on the radio’ briefly view that as a compliment…

Categories: Statistrickery, Technology Tags:

Serious violent crime map

July 28th, 2008 John B 1 comment

I’ve been busy working and being in pain (alcohol and seedy dives with steep slippery wooden staircases are not a good combination, but codeine and sodium naproxen are excellent painkillers); apologies for lack of posting.

Luckily for those of you who come here for evidence that we’re not all going to hell in a handcart having been murdered in our beds, Mark Easton at the BBC has been busy researching the data – and has an excellent post on levels of serious violence by geography, which also highlights the falling levels of serious violent crime.

Categories: Statistrickery Tags:

On why society is basically doomed

July 11th, 2008 John B 5 comments

Sunny has a good piece on Commentisfree on the knife panic – worth a read. Apparently the original version linked to this post here, but the Guardian subs took it out – bad Guardian subs.

Unfortunately, the comments indicate why we’re never going to get anywhere with the criminal justice system, other than locking up more people for being black and scary with no discernible benefit to anyone – the complete and utter cognitive dissonance of the public on the issue.

This is a shining example:

On the issue of violent crime per se, the statement that Sunny made that ‘overall violent crime has decreased by 41% since a peak in 1995′ is not believable.

Sunny, never understimate ‘common sense’. People have a fairly good idea of what is going on around them ….

Nonsesne. Common sense is the most overrated tool in the box and the cause of most stupid actions; and people have absolutely fuck all idea of what is going on around them. If you trust some random members of the public over the wider body of academic research on any issue, then you’re almost certainly wrong and definitely an idiot.

Unfortunately, most people on most issues trust random members of the public over the wider body of academic research, because most people aren’t very bright. Never mind that the conclusions it leads them to are entirely wrong; at least it saves them having to think…

Categories: Bit of politics, Statistrickery Tags:

That inflation thing again

June 25th, 2008 John B 6 comments

My personal inflation rate (RPI) is 3.5% to May 2008. What’s yours?

Categories: Statistrickery Tags:

An open letter to people who bought houses in the last two years

June 21st, 2008 John B 4 comments

I’ve been living in rented accommodation for the last two years despite being able to afford a house/flat, because it was incredibly obvious that house prices were going to peak, interest rates rise, and the economy slow down.

The fact that you did buy a house means that either you felt financially secure enough that even if all the above happened, you’d still be able to pay the bills (I know some people who’ve gone down this route; good luck to them), or that you thought buying a house was a supercool one-way ticket to infinite money at the expense of, erm, someone else.

If you’re in the first group, good luck to you. If you’re in the second group, then I’ve got to admit that I don’t find your financial travails spectacularly moving.

Relatedly, please can clueless muppeteers stop whining that “ooh, inflation is actually really high”? The CPI and RPI, both of which are still below 5%, are weighted by people’s actual spending on the actual goods included, so the fact that chips cost 15% more than they did is not evidence in favour of your proposition. Here is the actual breakdown of price changes by category (PDF). Read, digest, then put up or shut up.

Also, read this. It is true.

Categories: Financial arcana, Statistrickery Tags:

Sky in ‘not falling’ shock

April 15th, 2008 John B 13 comments

Here is a table showing projected growth in real GDP per capita [*] for 2007-08, according to IMF data released this month:

2007-08 2008-09 2009-10
Canada 0.0% 0.6% 2.0%
France 0.8% 0.7% 2.0%
Germany 1.5% 1.1% 1.9%
Italy -0.1% -0.1% 0.4%
Japan 1.4% 1.6% 1.8%
United Kingdom 1.3% 1.4% 2.1%
United States -0.5% -0.4% 1.9%

In other words, credible independent people who understand economic forecasting (note: not people who couldn’t tell GDP from a CDR but who think the PM is dour and boring and that therefore everything is going to the dogs) believe that the UK economy is unlikely to see recession, and is highly likely to outperform all of the G7 economies except for Japan over the next three years. Most other economic forecasters hold similar views, with a strong consensus that 2008 real UK economic growth will be 1-1.5%.

So why the hell is there such an insanely overhyped climate of doom going on, with people who shouldn’t really know better everywhere pontificating about how the government has ruined everything and left us all in a terrible position? Is it just that everyone who bothers going on about this kind of thing is a miserable get, whereas the people who think everything will probably be fine would sooner talk about Eurovision and football?

It’s partly due to misleading reporting, of course. Hands up who read this morning’s paper and came away with the take-out that UK retail sales fell when comparing March 2008 to March 2007? Wrong: they rose by 1.1%. The fall was in ‘like-for-like’ sales – i.e. new shops are opening faster than people are increasing their spending. That’s not great news if you’re a retailer, for sure – but it also for sure doesn’t mean that sales are falling…

[*] i.e. stripping out the effects of inflation, so you can’t whine about price rises, and done consistently on an international basis, so you can’t whine about CPI vs RPI. Oh, and please don’t slate the methodology the IMF uses to calculate GDP deflators, at least unless you have at least a master’s degree in a numerate discipline with some connection to economics.

Categories: Financial arcana, Statistrickery Tags:

Switch to our mobile site