Cunliffe slams English; English misleads House

Written By: Marty G - Date published: 6:14 pm, February 25th, 2010 - 44 comments
Categories: Economy, Parliament - Tags: bill english, David Cunliffe, GDP

Great to see David Cunliffe confront Bill English. After Cunliffe pointed out that the economy grew by an average of 3.2% a year under Labour compared to 2.6% in the 1990s under National and -2.2% last year under National, there was this question:

Cunliffe: how does he reconcile his statement that between 2005 and 2008 “the economy grew by less than 1% a year” with official statistics that show the economy grew at almost twice the rate he claimed?

English: I think he’s retailing a story from a couple of weeks ago on one of those left-wing blogs. And the problem there is that he is using nominal growth in the economy when the standard measure is real growth

Well, nice to see you’ve noticed my work, Bill, but you’re lying and you know it.

The numbers I have used (and the ones Labour must have used) are inflation-adjusted – they come from table 2.1 of the Stats NZ GDP release. Contrary to what English told the House, the 1.7% is real growth – ie. 1.7% over and above inflation. (click here to see my calcs on the right in full size and what would happen if you used nominal figures, from table 4.1 instead, you get 5.4% on average).

See, it’s actually English that got it wrong. To get 0.9%, he compared September Quarter to September Quarter, not September Year to September Year.

Why is comparing quarter to quarter wrong? Well, say you want to compare daily river flows. Do you compare the flow of the river between 6 and midnight, claiming it tells you the difference in the amount of water that flowed down each day? No, you compare flow over the whole days. If you want to compare annual GDP you compare GDP over the whole years, not just isolated quarters.

English knows by now that he has made a mistake/been lying. He knows that I’ve got it right and Labour has got it right. He was misleading the House in claiming that the dicrepency between Labour’s numbers and his was due to using nominal numbers. The real reason for the difference was his use of the wrong numbers, which gave him the wrong result. The fault is his, and he is trying to mislead the House to cover for it.

Final words to Cunliffe:

Is the Minister aware that over the 9 years of the Labour Government GDP growth averaged 3.2 percent, which is higher than under the previous National Government and obviously his own; and is he further aware that that strong and sustained expansion was achieved at the same time that net debt was cut to zero, gross debt was cut in half, unemployment was less than half the current rate, thousands of New Zealanders were lifted out of poverty, and the minimum and average wages rose every year; if so, why does he not just admit to New Zealanders what is patently obvious: that he is shonky in his use of figures, and he has no plan for growth?

44 comments on “Cunliffe slams English; English misleads House ”

Comments are now closed

BLiP 1

25 February 2010 at 6:43 pm

Listening to the rhetoric in the lead up to and then the closing speeches on the ACC privatisation legislation and now to the deliberate obfuscation of the GDP figures, anyone with any confidence in National Ltdâ„¢’s ability with figures is either an adviser hoping for some sort of pay off or deliberately fooling themselves.

Well done Marty – and belated Onya to Cunliffe.
Clarke 2

25 February 2010 at 6:49 pm

Cunliffe has beautifully underlined the fact that English has basic numeracy problems.
- coolas 2.1
  
  25 February 2010 at 6:58 pm
  
  Anne Tolley should give him a National Standards numeracy test – but I forgot – she doesn’t understand what they are.
  - Mr Magoo 2.1.1
    
    25 February 2010 at 7:31 pm
    
    Now that is a good one.
    
    I wonder what numeracy age old billy boy would come out with?
    
    10, 15?
gomango 3

25 February 2010 at 7:02 pm

No marty – you have got it wrong with statistics (yet again).

See, it’s actually English that got it wrong. To get 0.9%, he compared September Quarter to September Quarter, not September Year to September Year. No, you are completely wrong..

The percentage change from the previous quarter ie for September is the GDP in Q3 divided by GDP in Q2 – 1. Raw data from table 2.1. You can replicate his calcs in about 5 seconds, average the 12 relevant quarters in th lower half of the table and multiply by 4.

Two figures are bing bandied out – the English ones as I have described (clearly not as you describe). And your river flow analogy is completely wrong as neither statistics quoted have a “flow gap” as your analogy does.

The other one quoted (a la Cunliffe) are the year on year figures. Here for September for instance they take the total GDP over the last 4 quarters and divide it by the GDP total over the 4 quarters prior to that.

So the two numbers look at very different things – percentage growth change in the last quarter compared to the previous quarter, and growth change in the last 4 quarters compared to the 4 quarters prior to that. C’mon Marty, catch up, this is barely even economics 101.

Whose figures are right? They both are, depending on whether you use the annual series or the quarterly series. What is wrong is your childlike statistical analysis which would fail NCEA level 1.

And as an aside, question time drove me nuts when National were in opposition, and it drives me nuts with Labour. Why do they frame their questions so poorly? So much grandstanding results in a complex question that the minister can provide a smart arse, non-answer to. Look at Cunliffes final question above – yes it makes him feel smug in the delivery but does it achieve anything apart form making cunliffe look smarmy? National were guilty of exactly the same thing when in opposition. It just makes the questioner look like a dick to the mainstream public.
- BLiP 3.1
  
  25 February 2010 at 7:29 pm
  
  Economics 101 – a la Mankiw who writes in his textbook that changes to the minimum wage will effect unemployment? Mankiw who stood over Dubya’s Council of Economic Advisers cheer leading for Greenspan?
- sk 3.2
  
  25 February 2010 at 7:30 pm
  
  You are an idiot. If you average the quarterly annualised numbers for the three years to Sept 2008, you end up with 4.1%. But that is the wrong way to think about it. Averaged quarterly numbers are meaningless
- Marty G 3.3
  
  25 February 2010 at 10:01 pm
  
  gomango, what the hell are you smoking?
  
  “The percentage change from the previous quarter ie for September is the GDP in Q3 divided by GDP in Q2 1. Raw data from table 2.1. You can replicate his calcs in about 5 seconds, average the 12 relevant quarters in th lower half of the table and multiply by 4”
  
  If you are working annual growth you add up all the quarters for one year, all the quarters for the next, divide the second by the first, take away 1, and times the result by 100. That is the annual growth figure, that is what English is claiming he is talking about, that is what any one would be talking about when talking about annual growth.
  
  On table 2.1 they have march year figures, you can replicate them by adding the four quarter of each march year (ie for the March 2009 year you add together June Quarter 08, Sept Quarter 08, Dec Quarter 08 and March Quarter 09 – 33532+33242+32960+33036 = 133494, the March year total provided by stats = 133485, the 0.007% difference is due to rounding).
  
  What English has done is just take each sept qaurter divide it by the previous one minus 1 times 100. That does give you the sept quarter on sept quarter growth but it doesn’t give you the annual growth rate. just like measuring 6 hours a day of a river’s flow and comparing it does not give you the same result as measuring the whole day’s flow and comparing it.
  
  You most certainly do not average the growth of all the quarters. That makes no sense in the slightest.
gomango 4

25 February 2010 at 7:03 pm

coolas – send one of those tests to marty too……….
- coolas 4.1
  
  25 February 2010 at 7:23 pm
  
  ok – for Marty
  1) how many bloggers does it take to show a Finance Minister is in-numerate?
  2) who was the square root of Richard Worth?
  3) if 1 is for dumb & 10 is for brilliant where does Tolley score?
sk 5

25 February 2010 at 7:25 pm

What the hell are you going on about? Arithmetically, the quarterly numbers sum to the annual. WTF? If you are talking economic performance the correct numbers are the annual. So what is childlike about Marty’s analysis?

By the way, if you looks at the Stats numbers 0.9% does not come into, whichever way you slice it.
- sk 5.1
  
  25 February 2010 at 7:27 pm
  
  That question was for gomango
mickysavage 6

25 February 2010 at 7:55 pm

Like everything else about this Government Blinglish’s figures are pure spin and have nothing to do with reality.

Blinglish has done it for a while. His “decade of lost opportunities” rant really grates because it is not true. Out of Parliament he acknowledges that Cullen did a sterling job but in Parliament lies for political purposes.

I thought that MPs who lied to Parliament could be hauled before the privileges committee?
gomango 7

25 February 2010 at 7:57 pm

sk – i’ve explained very clearly, shall i use words of one syllable for you?

Go to table 2.1

Sum cells R36:R47

Divide by 3

Got it? (Sorry, some of those words had 2 syllables).

You don’t understand how GDP numbers are presented. Arithmetically the quarterly numbers don’t add to the annual. I explained once before., I’ll explain again.

To get the annual numbers you add the GDP for the last 4 quarters, and divide by the sum of the GDP from the previous 4 minus 1.

The quarterly growth rates are this quarters GDP divided by previous quarters GDP minus 1.

The quarterly numbers don’t add to the annual. Do you want to go thru the algebra? I’m not sure how i can explain it more simply.

I suggest you do some research before wading into a debate above your pay level. Do you get it now?
- sk 7.1
  
  25 February 2010 at 8:11 pm
  
  Gomango, you are quite the joker aren’t you? Do you spend your evening gomangling data that you do not have the brain to analyse? Column R is the quarter on quarter changes. To get an an annualised growth rate you have to quarter divided by quarter (-1) X FOUR!!!!
  
  I really really hope that Blinglish did not make that same mistake, else we are all completely screwed
- sk 7.2
  
  25 February 2010 at 8:18 pm
  
  also, no one takes (q3:2008 / q3:2005 -1)/3. That is not the way to look at GDP numbers – which after all are a volatile series. Point-to-point over an extended period can generate weird outcomes. I am surprised anyone in English’s office would present data that way. But then again, that’s Treasury for you
  
  Oh, and annual GDP is the sum of the quarters.
vto 8

25 February 2010 at 8:08 pm

sheesh, talk about statistics and statistics.

The real issue is what caused that period of growth while labour just happened to be in power. That growth had NOTHING to do with labour. You all know it. You’re just as bad as English with your twistings.

Oh, other than the other real issue of obfuscation by politicians. Quite honestly none of them can be relied on for the full picture. English has “stand clear – dodge at work” imprinted all over his forehead. Cunliffe has “I’ve got a shinier tie” all over his.

Having spouted off though I do have to agree that English has got away with an awful lot of bullshit recently. I have listened to him from time to time and he seems to virtually make it up as he goes. He seems a bit like Winston Peters, just keep firing off and keep moving and nobody will keep up. Except when they do finally catch up a-la Owen Glenn styles …
gomango 9

25 February 2010 at 8:11 pm

vto. Stop talking commonsense. You are spot on with every comment.
Chris 10

25 February 2010 at 8:18 pm

I’ve had experience of English and frankly I wouldn’t trust the guy as far as I could throw him. I suspect many in the National caucus feel the same way.
gomango 11

25 February 2010 at 8:19 pm

sk. sigh.

The way GDP annualised rates are presented by the stats dept are as I described, not as you have described. And in any case in your method you wouldn’t multiply by 4, you’d raise to the power of 4.

The way you described is essentially what bill english is talking up averaging the quarterly growth rates. He thanks you for your defense. You are a moron.
- sk 11.1
  
  25 February 2010 at 8:22 pm
  
  The Stats dept reports a quarterly GDP (which is activity that quarter), and compares it to the previous quarter, and to the year before.
  
  What English and you have done is very different, and very arbitary. . . Sigh
gomango 12

25 February 2010 at 8:35 pm

no sk. yet again you are wrong.

Stats dept quotes two figures.

Quarterly = this quarters gdp divided by previous quarters gdp -1.

Annual = sum of last 4 quarters gdp divided by sum of the previous 4 quarters gdp -1.

Go to the stats website or the rbnz website or the treasury website and educate yourself. I learned this in high school.

This is specifically what the Stats department does NOT do:
To get an an annualised growth rate you have to quarter divided by quarter (-1) X FOUR!!!!
- sk 12.1
  
  25 February 2010 at 8:44 pm
  
  Gomangling, it took me a while to work out what you and Mr English were going on about, as it was so unorthodox. I am glad you studied this in high school. Good for you.
  
  The point is that a quarterly GDP number tells us about what the economy has produced in any one quarter. To get a true sense of how it is performing you have to look at the annual numbers, or take some average of the quarterly annualised numbers.
  
  This is not what you or Mr English did. You both took (Q3:08/Q3:05-1)/3. That may be fine for you with your high school education, but it is completely unacceptable for a finance minister to present his country’s economic performance in that way.
  
  Oh, and my GDP spreadsheet goes back to 1960. Maybe you want to educate yourself a bit more
lprent 13

25 February 2010 at 8:39 pm

Good to see David score points… We need it..

Now about that question to John Key if he has confidence in his ministers?
gomango 14

25 February 2010 at 8:53 pm

You’re completely (wilfully?) missing the point……..

what Bill English is doing is exactly what you say is the right thing – averaging the quarterly numbers!

here are the quarterly numbers…..

-0.2%
0.3%
0.0%
0.1%
0.3%
1.3%
0.8%
0.7%
1.0%
-0.4%
-0.6%
-0.7%

Lets do what bill english did. Average them, multiply by 4. Answer 0.9% per annum.

Lets do what you say.
Take those same numbers multiply each by 4.

-0.7%
1.4%
0.0%
0.3%
1.3%
5.3%
3.2%
2.8%
4.0%
-1.5%
-2.5%
-2.7%

average them. Answer = 0.9%per annum.

Question: are you Bill English?

The correct way is the way the Stats dept does it.

(Sum of last 4 quarters) / (sum of 4 quarters prior to that) – 1.

Please keep digging Bill sorry i mean sk
- NickS 14.1
  
  25 February 2010 at 9:07 pm
  
  The stupid, it burns.
  
  Per annum = Per year
  
  So doing it quarterly only gives you the quarterly GDP growth, to get the per annum GDP growth rate you need to sum up the whole year’s GDP…
  
  This is so utterly, completely, ridiculously simple, you can but only be an utter morosoph, an mathematically illiterate morosoph.
  
  Lawl.
sk 15

25 February 2010 at 9:04 pm

Gomangling, you have made it very clear how Mr English calculated the numbers. Great. I couldn;t make sense of his 0.9% number. Thank you. But as I said, that is no way for a Minister of Finance to present his country’s growth performance.

IT IS COMPLETELY WRONG!!!!

To calculate economic performance you have to take annual averages. No ifs, no buts. Bill English is out of line to twist GDP numbers like that. After all, artificially talking down our LT economic performance is unacceptable. No one looks at GDP numbers the way you – and Bill – have presented them
gomango 16

25 February 2010 at 9:08 pm

I see you have surrendered sk, well done.

Surprised you couldnt make sense of the 0.9% number as you actually provided the formula above and suggested it was the right way to look at GDP……….. I guess you’re backing off that statement now? Tip- avoid saying “annual averages” thats not how its done – refer to my previous comment yet again.

I’d actually go with the Stats dept rather than your ramblings.

[lprent: perilously close to a pwned statement. I’ll let it ride for the moment. ]
- sk 16.1
  
  25 February 2010 at 9:14 pm
  
  What a jerk. Of course I provided the formula. But only Bill and you would consider it useful. The Right is so determined to expunge Labour’s record that they will distort numbers, and present them in a way no one else on the planet would consider doing.
  
  NZ’s growth was 1.7% in the three years to Sept 2008. Marty’s original point stands. All you have proved is that the number’s English presented in the house were distorted to make a political point, which is unacceptable for a Minister of Finance.
- Marty G 16.2
  
  25 February 2010 at 10:10 pm
  
  oh dear god. I only got to the rest of this stuff from gomango now.
  
  Honestly dude, you don’t have the slightest clue what you’re talking about. Stats works out annual rates exactly as I have described but they only do it for March years.
  
  See how there’s no Paul Walker or any of the economists coming to even dispute this with me? See how David Farrar isn’t arguing the numbers with me? That’s because this is basic basic stuff and I am completely right.
  
  And don’t think for a second that Labour took my word on this. They would have gotten their economics people to do it.
  
  Hell, you know what? I was so surprised when I initially heard this claim from English that after I got the results, I checked and rechecked them and then just because I couldn’t quite believe the minister for finance was making this kind of elementary error, I rang up a financial wizz whose a mate of mine and asked him, without any prior indication of what I wanted it for, to calculate the annual growth rates for September years 05/06 to 08/09 , and he came back with exactly the same numbers as I did.
  
  I don’t know what you do for a living gomango but I saw as hell hope you don’t have any part in reporting annual performance of a company.
  - Lanthanide 16.2.1
    
    25 February 2010 at 10:37 pm
    
    Marty, I’m curious, are you in a stats-related field? You post all of these graphs and stats all the time, which I’m guessing is because you have a passion for that sort of thing.
    - Marty G 16.2.1.1
      
      25 February 2010 at 11:40 pm
      
      yeah, stats-related field.
      
      But i’ve always been a bit of a geek for this kind of thing. Probably one of the reasons they asked me to join, blend of skills. Wish I could write like Irish though.
gomango 17

25 February 2010 at 11:12 pm

ummmmm….. actually no marty

Top of page 2 here: http://www.stats.govt.nz/~/media/Statistics/Browse%20for%20stats/GrossDomesticProduct/HOTPSep09qtr/GrossDomesticProductSep09qtrHOTP.ashx (annual figures on a quarterly basis, not just march quarters)
Calculation performed as follows from the famous table 2.1 in the GDP report.

ie (sum of 2008Q4 thru 2009Q3) / (sum of 2007Q4 thru 2008Q3) etc

I think we’ve got way off the point – point I was making is the numbers Bill English is quoting actually are a legitimate way to describe growth. You wont win an argument calling him a liar for using them.

Again here are the 12 quarters he is talking about:

-0.2
0.3
0.0
0.1
0.3
1.3
0.8
0.7
1.0
-0.4
-0.6
-0.7

Average them – you get 0.9% per annum, or compound and annualise them you get 0.9% per annum.

The annual growth rates as quoted are not equivalent to the quarterly rates annualised. They are quoted on a different convention.

The reason why Cunliffe is quoting Annual march figures is they conveniently don’t go negative until March 2009, ie “Nats mismanage the economy”

The reason why English is quoting an average of the quarterly figures (just as legitimate) is because they conveniently show growth first went negative in the March 08 quarter, ie “Labour squander the good times”.
- Marty G 17.1
  
  25 February 2010 at 11:16 pm
  
  gomango. You do not average quarterly growth figures to get annual growth figures. English said “annual growth averaged less than 1%”.
  
  You do not average quarterly growth figures to get annual growth figures.
  
  You add up the years’ quarters, then calculate the % difference between the years.
  - BLiP 17.1.1
    
    25 February 2010 at 11:23 pm
    
    Oh – Marty, you’re spoiling the fun! gomango was making such a twat of himself with his “Economics 101” I thought we could have got a few more lolz!!
    - Marty G 17.1.1.1
      
      25 February 2010 at 11:38 pm
      
      sorry BLiP 🙂
      
      Doesn’t he realise that none of the righties are agreeing with him, they’re just too embarrassed to tell him what he’s doing wrong?
Marty G 18

25 February 2010 at 11:29 pm

I want you to try this with me in excel:

imagine 15 year old person and you want to know the average annual increase in the amount of food they ate over the preceding five years of their life.

Create the food intake for the five years, random amounts but generally rising if you like.. Now to calculate the average growth it’s simple, you just compare each year to the preceding one and average the growth rates: (B/A-1)*100 gives you the % B is greater than A.

But let’s make it harder and break each of those five years into uneven quarters, the quarters for each year still totaling the same as the previous annual totals create 20 quarters of food intake.

By your method, we would look at what they ate each quarter, get the % growth (or decline) for each quarter over that time period, add up the quarters and divide by five. Do it. Notice how the number comes out completely different to when we averaged those years at the start.

But now try adding the quarters for each year and then calculating the % change between each year and the preceding year, then averaging that %. Effectively, we’re putting the quarters back into years then calculating the average annual growth. Hey presto, same % as the first calc we did.

I want you to actually do this or show why it’s a flawed methodology for calculating the average growth rate
SPC 19

26 February 2010 at 12:25 am

Maybe gomango works as a climate change denialist.
wtl 20

26 February 2010 at 12:25 am

Okay, now I see that two different methods give the 0.9% figure. Both are flawed, but for different reasons and to different extents:

1) The gomango magic average the quarter’s to get the yearly figure. Obviously, this is completely mathematically unsound.

2) The Sep Quarter to Sep Quarter method suggested by Marty. This is inaccurate because it only takes into account one quarter of the year, and misses out the rest of the year. Only mathematically sound if you were only interested in the Sep quarters (for some reason). And obviously it is very dodgy if you deliberately chose the Sep quarters because they gave you the answer you wanted (i.e. poor growth).

Which one did English use? Surely not (1), right?
- Marty G 20.1
  
  26 February 2010 at 12:34 am
  
  well, (2) at least could look to the lay person like a legit way to measure it, whereas (1) is batshit crazy.
  
  Actually, if you look at the first post on this http://www.thestandard.org.nz/our-innumerate-finance-minister-friends , it was (1) that I thought of first and discovered it came out at 0.9% but then someone pointed out (2) also comes out at 0.9% and I’ve been assuming they got there with (2) ever since.
  
  One does half wonder, given the passion of his lone defence of the insane, if gomango might be the person in English’s office who started this whole saga off though….
BLiP 21

26 February 2010 at 1:04 am

The reason why English is quoting an average of the quarterly figures (just as legitimate) is because they conveniently show growth first went negative in the March 08 quarter, ie “Labour squander the good times’.

Yet another National Ltdâ„¢ supporter calling Blinglish a big fat liar . . . hmmmm – maybe he is!!
Slartibartfast 22

26 February 2010 at 11:32 am

confounded Bistromath !
Descendant Of Smith 23

27 February 2010 at 8:57 am

I read these calculations several times and still don’t get it. I’ll keep stretching my brain to understand however.

I was always taught that you can’t average percentages to get any sort of accurate result. The exception is where the denominator of all the figures you are using is the same – such as test scores out of 100.

In this instance as the quarterly growth is measure is measured from the quarter before the base denominator changes each quarter so you can’t average the percentages.

By way of illustration:

1st quarter Starting point 100 + 10% growth = 110
2nd quarter Starting point 110 + 10% growth = 121

The growth for the six months is 21% not 20%.
Now I can see how you could say that the average quarterly growth is 10% but would it be more or less correct to say the average quarterly growth for that six months was 10.5%.

If A is correct then would Bill English’s statement be more correct by saying that the average annual growth was 0.9% per quarter.