drplokta: (Default)
[personal profile] drplokta
Thanks to The Register, here is a nice graph of the decline of manufacturing in the UK since 1945. You'll notice one small problem with the "decline" bit.

(no subject)

Date: 2010-02-22 02:40 pm (UTC)
andrewducker: (Default)
From: [personal profile] andrewducker
So far as I can tell, that doesn't compensate for inflation.

Inflation from 1940 is about 4700%, so I'd expect the values to go from 40 to 2000, not 100.

(no subject)

Date: 2010-02-22 02:44 pm (UTC)
From: [identity profile] coth.livejournal.com
It said it was a 2005 index in the article.

And is also fairly explicit that a computer programmer working on the manufacture of aircraft engines is counted towards the value of manufacturing as an industry.

(no subject)

Date: 2010-02-22 02:48 pm (UTC)
andrewducker: (Default)
From: [personal profile] andrewducker
Sorry, I meant to say that I'd expect the 1940 value to be 8 rather than 40, if it's not taking inflation into account. I got it the wrong way round in the translation from brain to fingers.

Without knowing whether the figures do take inflation into account or not it's very hard to tell if they're meaningful or not.
Edited Date: 2010-02-22 02:48 pm (UTC)

Inflation

Date: 2010-02-22 02:59 pm (UTC)
From: [identity profile] armb.livejournal.com
http://www.statistics.gov.uk/CCI/nugget.asp?ID=169 says "The index is measured at base year prices (currently 2000)."

Re: Inflation

Date: 2010-02-22 03:04 pm (UTC)
andrewducker: (Default)
From: [personal profile] andrewducker
Aaaah, so any items that are sold are weighted at the price that those items would have sold for in 2000? Must be tricky for electronics, but it does make sense.

Ok, that makes me a lot happier. Thanks!

Re: Inflation

Date: 2010-02-22 03:17 pm (UTC)
From: [identity profile] nmg.livejournal.com
Maybe - it's a good point.

(no subject)

Date: 2010-02-22 03:15 pm (UTC)
From: [identity profile] nmg.livejournal.com
Really? Gosh. You'd better phone the statisticians and tell them that they've overlooked this crucial yet trivial point! ;)

(the Reg forums are full of eejits making this observation as if no-one else had thought of it)

(no subject)

Date: 2010-02-22 03:19 pm (UTC)
andrewducker: (Default)
From: [personal profile] andrewducker
Que?

That graph would mean two different things depending on what it shows. If it shows "cost of goods produced at the time" then it's showing a massive decline in manufacturing. If it shows "cost of goods produce, normalised to a standard date" then it shows a decent improvement in manufacturing over time. All I wanted to know is which one I was looking at.

(no subject)

Date: 2010-02-22 03:35 pm (UTC)
From: [identity profile] nmg.livejournal.com
The *default* assumption to make with datasets like this is that prices are normalised to a base year; the data isn't really any use without this.

(no subject)

Date: 2010-02-22 03:37 pm (UTC)
andrewducker: (Default)
From: [personal profile] andrewducker
I don't like making assumptions about data visulations, particularly not with ones that start somewhere like The Register.

December 2016

S M T W T F S
    123
45678910
11121314151617
18192021222324
2526 2728293031

Most Popular Tags

Page Summary

Style Credit

Expand Cut Tags

No cut tags