Friday, December 30, 2011

This Is NOT How You Calculate Inflation (UPDATED)

I wrote about DS Anwar’s speech regarding the ETP a week ago, but my comments were based entirely on the coverage by the Malaysian Insider, and not the speech itself. I stumbled on that via JMD’s blog.

Just so you know, PEMANDU have issued a rebuttal (without naming names).

But what caught my eye in the original speech was this paragraph that wasn’t mentioned in the news coverage:

The average inflation between 2001 and 2005 is 4.8%, reflecting the first energy price shock of 2003 that saw average crude oil prices moving beyond the psychological US$30 per barrel mark. The average inflation between 2005 and 2009 is even higher at 6.6% as a result of the 2008 crude oil price rally that saw the energy prices sky-rocketing throughout the world.

The inflation numbers sounded wrong to me, so I looked back at the actual data. Honestly, when I figured out how he got the inflation numbers, I had to laugh.

Here’s how they did it:

Somebody took either the monthly or annual CPI index numbers from 2001 to 2005, and averaged them, which gives an answer of 104.8. Divide that by the base index number (2000=100), and you get a 4.8% increase.

For the second period, they did the same thing, only using the rebased 2005 index with numbers from 2005-2009 which gives 106.55 (monthly). Divide that by 100 and you get a 6.6% increase.

The problem here is that doing it this way doesn’t give you average inflation – it gives you the average price level. And inflation is the change in the price level, not the price level itself.

So for the record, this is how it’s supposed to be done (for you finance types, you’re actually looking for the compound annual growth rate – CAGR – in the index numbers). Assuming we use the annual index (which averages the monthly index numbers):

For 2001-2005:

5√(2005 index number)/(2000 index  number)

=5√(109.1/100)

=1.0174

In other words, average inflation from 2001 to 2005 is 1.74%.

 

For 2005-2009:

5√(2009 index number)/(2005 index number)

=5√(112.1/100)

=1.0231

…which is just 2.3% per annum.

That’s a far cry from 4.8% and 6.6%. I think Dato’ Seri Anwar needs a new research assistant –preferably someone who knows math.

(H/T Jebat Must Die)

UPDATE:

Seems I need math lessons myself. The second equation  should be:

4√(2009 index number)/(2005 index number)

=4√(112.1/100)

=1.02899

Which gives 2.9% average inflation from 2006-2009.

The difficulty arises because the article mentioned 2005-2009, which is of course a little misleading, because if you’re taking 2005 as the base year inflation would be zero in year 1 (2005 prices over 2005 prices means zero increase).

To get around this, I recalculated using base 2000=100:

5√(2009 index number)/(2004 index number)

=5√(122.21/105.86)

=1.02914

…which is slightly higher, but still around 2.9% on average for 2005-2009.

Thanks to the anonymous poster for pointing this out.

37 comments:

  1. Anwar the know all things will always be like that.

    And u should include the 'know all things prof' and 'i jilat all your things fan'...

    ReplyDelete
  2. From 2005-2009 index,would it simply mean that the basket worth RM100 will in 2009 cost RM112 i.e 12% more?
    This seems so farfetched for most people and figures between 30 to 50% seems more likely.
    What is the increase for the same period for your "pain index"?

    Definitely for us in low income the pain is more than the middle class.Disposable income n thus savings are squeezed to meet day to day expenditures.Is there any mechanism to have CPIs based on different income groups aka your pain index?If you ask most non experts ,they would likely believe 6% rather than 2.8% cos they fill the pinch n its a hard painful one.

    If govt continues to rely on official CPI blindly,policies will not impact the target group positively.

    ReplyDelete
  3. @anonymous,

    Exactly right - prices in 2009 were on average 12% higher than in 2005.

    The reason for the seeming dichotomy between increases in actual living costs and the official numbers is a couple of reasons - one is, as you note, inflation is different for different income classes (something I've commented on before). It's more than possible that the inflation rate for those in the lower income brackets is running at double the official overall rate.

    The other is that while prices increased spectacularly in early 2008, they also dropped sharply in late 2008. As a result, the inflation data doesn't really "show" the increase in prices as much as you would think. There's a huge bulge in the index numbers in the middle of 2008, but you'd never know it looking at the year end figures (or the averages). The average for the Pain index (2005 prices) for 2008 was 118.4 and 117.7 for 2009 (peak was 125.1 in August 2008).

    One other factor is that with the recession, wages froze in many industries which probably contributed to the perception of higher inflation - real wages likely fell in 2008/2009.

    With respect to different inflation indexes for different income levels, I've advocated that before (see the link above). The mechanisms are already available, as DOS collects data on both household expenditure and income, which means calculating different consumption baskets ought to be a trivial exercise.

    If even Kenya can do it, why can't we?

    ReplyDelete
  4. bila baca blog hishamh org kg mcm saya ni lambat nk paham...

    ReplyDelete
  5. actually who really care if CPI is 1.75% or 6.6% when the basic living expense increase by far more than that each year?

    ReplyDelete
  6. Funny you should say that - mine doesn't feel like it's increased much for the past few years. And I do the grocery shopping for my household.

    The relevance of the inflation number is that the implication for monetary policy. If indeed we had average inflation at 6.6%, that should have triggered an increase in interest rates much higher than we actually saw - by rights, the OPR should have been at 7% and BLR AT 10%-11%.

    ReplyDelete
  7. enough with this bruhaha..... people dont give a damn on this cpi thing... all that they care.... price hiking... income stagnant... all that crap is just a numbers, numbers is bullshit... people is real, living in a real world... and they are suffering....

    ReplyDelete
  8. I think people are doing pretty well. Even my general workers are carrying IPhone and I am paying them average wages ;-)

    ReplyDelete
  9. en. hisham..rasanya bagi pengiraan diatas kena guna punca kuasa 4 instead of 5..can u double check on that..

    ReplyDelete
  10. Dang, you're right - but I believe it's just for the second period (2005-2009).

    Post updated with recalculated figures.

    ReplyDelete
  11. this is great, thanks. numbers are so dangerous, for someone yang clueless about this kinds of things like me, easy to be misled.

    but honestly, I would like to share, me being a young private sector worker, I do not find the prices and taxes in Malaysia a burden, nor do I find my wages as being not enough. I can afford groceries for my family, pay the bills at my house, pay the bills for my mom's house, give her money for groceries, contribute to pay my sick brother's monthly medical bills and also take my family and my mom on the occasional overseas holiday (thanks to budget airlines and forward planning) and shopping sessions. we drive an average national car and I can also pay for my mom's average national car. we live in a nice but small house within my means and also manage to save every month. and my mom, living on her teachers pension can survive quite comfortably.

    I live in the real world too, and I think I do okay, just living within my means.

    ReplyDelete
  12. in fact he dont need to be good. He needs a team like Pemandu to do it for him. Dont tell me u're from Pemandu, which only know how to calculate real figure and tell another figure to media.

    ReplyDelete
  13. Sally, thanks for your contribution. I think for a small segment of the population, inflation doesn't seem like much - I'm certainly included in that category.

    Just bear in mind that less than 10% of Malaysian households (not individuals) earn more than RM5000 ringgit a month.

    @Khai Soon

    No, I'm not with PEMANDU :)

    To be fair to PEMANDU, most of the numbers they are dealing with were generated via the New Economic Model framework, which consisted of some fairly decent local economic thinkers (which doesn't make them necessarily right of course).

    In that sense, DSAI is right - it depends on the assumptions you take (and I've had a dig at them myself).

    But I would've thought that an aspiring PM-in-waiting would have somebody checking his numbers. It's embarrassing to get them publicly wrong (and that includes me too!).

    ReplyDelete
  14. Pemandu figures is more of a deception..
    They are projecting RM 1.7 trillion GNI in 2020.And they are also claiming that they are creating 800 bil of that GNI thru their EPPs,BOs & multipliers.Thus the BAU GNI is only 900 bil.On real terms (with assumed 2.8% inflation) thats only 0.4% BAU GNI growth rate.

    U can't transform unless u understand the BAU i.e where we are now n going fwd pre-transformation.Thus its just another corridor exercise i.e draw a circle and claim thats my output.

    This is the bullshit that all of u economist shld expose.

    Realistically,BAU GNI growth of 6-8 % nominal is attainable.If there's need for Pemandu is to go beyond that BAU.

    Please don't tell me that increasing yield of FFB & OER are not BAU.Please don't tell me O&G marginal field exploitation etc are unheard of pre-Pemandu.

    Emperor got new clothes all over again.

    ReplyDelete
  15. @anon,

    If you read any of my early posts, I think you'll find I mostly agree with you. To me, most of the heavy lifting towards getting to the high income target will come through two channels - a lower demographic dependency ratio (i.e. higher ratio of workers to overall population), and the Penn effect (higher incomes=higher price level=currency appreciation).

    The labour force is currently increasing at a rate nearly three times faster than overall population growth; the Ringgit has been appreciating on average of nearly 3% per annum.

    ReplyDelete
  16. So,why isn't anybody telling all this to Najib?Pemandu's micromanaging will hv a negative effect.Investors are not fools.Other govt servants incl Ministers are beginning to complain.Its not a case of lembu punya susu only...but the arrogance to bulldoze bad projects.
    Pls tell Najib the real scenario before Pemandu destroy us as MAS was destroyed.

    ReplyDelete
  17. My ideas aren't orthodoxy, and you have to factor in that "expert" really means you're good in one very specific area, and might not have a broad enough depth of knowledge to appreciate the total picture. Call in the World Bank and they'll talk about development issues; McKinsey will push "change management" strategies.

    At this stage, there's too much political capital invested in the ETP/GTP for it to be given up. But I don't necessarily see it as all bad. The ETP projects get the headlines, but there's a lot of good work being done that doesn't involve private sector investment and mega projects e.g. universal pre-school education. And I take civil service complaints as a good sign - I hear lots of rice bowls are being upset.

    ReplyDelete
  18. Roy,
    Things here are not as bad as Spain or US. But general workers carrying iphones may more likely be caused change in cultural values. From traditional conservative savers to modern consumerism. This It is also likely that these general workers are merely soaking up the vast pool of second/third hand phones. For some who cannot afford to buy AP cars or terrace house in urban areas, owning a smartphone is probable the only form luxury they have.

    ReplyDelete
  19. But 0.4% BAU GNI growth is a big big misrepresentation.We got to start transformation with the TRUTH or at least a better approximation of one,

    ReplyDelete
  20. Sally,

    As hishamh pointed out less than 10% of Malaysians earn more than RM5000.It is rather hard to define the middle class.Maybe even harder than setting poverty line. Income range or job type alone might not be enough.

    Despite of all our complaints about things getting expensive, it did not really put a dent in our consumption.

    But I do believe there are some concerns:
    1)CPI may not have taken enough urban residential and prime commercial land price increases into account. Even in remote places like Bahau house prices had skyrocketed in recent years. Land cost is the biggest part of individual expenditure and major part of retail business.
    2)Control goods(poultry,rice,etc) and common goods(milo)are probably in line in with CPI. But for goods and services outside those classes, I believe they are above the reported CPI. Dining out is way more expensive. Medium range milk formula spiked across the board(from personal experience). Pringles had shrunk.
    3)Over the past decade, Malaysian wages roughly increased about 2% above CPI annually. But we probably need a clearer picture of how different sector wages measure against non official CPI expenditure (goods/services and land cost).

    Malaysians in general are still able to enjoy lots of material stuffs and maybe more than what they did a decade ago(ipad,high end dining etc).

    It would be inaccurate to say that Malaysians paid for this increase in consumption through increase in household debt. Because house debt was downtrending up to 2007. And then just lompat in 2008 and stayed there till now.

    That said, there is probably another thing that need more study. And that is measuring personal debt(mortgage,auto loan, student loan,credit card debt)against different age group and their income level plus job sectors.

    My worries is Malaysia having something like 'housing wealth effect'. Especially for those who had hire-puchased a house 5 to 10 years ago. Seeing that their houses increases in price might help in increasing/maintaining their consumption pattern or simply take on more debt that they would otherwise choose not too with their marginal wage growth.

    ReplyDelete
  21. @anonymous,

    That's what blogs like this and others are for.

    ReplyDelete
  22. You are doing a fantastic job breaking down esoteric stuff to simple nuggets we are at least able to nibble at.

    Re Mr Yang's comment>>lots of gap in info or myb we do not look hard enuff to find them.Thus ,we are sometimes guilty of making assumptions.
    Generally,think purchasing power is lower now than 30 years ago.However,temptation to purchase have increased with various consumer products,accessible food outlets,branded goods etc.Product cycles hv shortened with new models launched every so often.And a lot of service becomes necessities i.e Astro,mobile services,wifi,3G etc.
    I sometimes meet families earning RM3k/mth and discuss their finances.Believe me..its a struggle and debt fueled cashflow mgmt.And over last 2-3 years costs hv really appreciated.
    I really pray that govt should do more surveys/data collection and analysis of this low income group.Without that,there's no way a holistic solution can be structured.Afterall..in 2020 ETP hv "programmed" 8 mil to earn below RM 1.5k.Thats 50% of the workforce on the brink of poverty.

    That,s not and never will be sustainable.

    ReplyDelete
  23. Anon,
    Its pretty obvious that most can afford CPI stuff. So obviously we need to look at non CPI stuff.

    Debt agregates against GDP may not tell the whole picture. That is why I reiterate that we need to look at age group or demographic components in the study of debt vs income level.

    There are probably many examples of Malaysian in their 50s owning 2 or 3 houses in urban area. While their early 30s counterpart struggling with their first mortgage.

    Excuse for being paranoid here, I m very worried about land price being the elephant in price inflation basket.

    ReplyDelete
  24. Oi Mun,

    With reference to your earlier comment (3.41):

    1. Land costs and housing purchases are not part of the CPI (these are assets, not consumption items). What they have in the CPI instead is rents and imputed rents, which have not kept pace with house price increases (ask any real estate broker). Hence the perception that prices have increased, but the CPI has not.

    2. Only partly true, as food away from home is included in the CPI (look at the table at the bottom of the linked page). I don't think this covers trips to fancy restaurants though. That might be covered under the restaurants and hotels sub-index, which is also part of the CPI. Both this and food away from home are generally increasing faster than food at home.

    Two things to note here - the weightings attached to each food category, and the incredible volatility of food prices (read the second to last two paragraphs in the link). What I think is going on is that people are noticing when prices of food items shoot up, but don't take note when they drop. That's in line with empirical studies on human behaviour where bad news is given more weight (sticks in the mind) than good news.

    A second thing is that they might equate food price increases in one item as applying to prices of other food and items in general, despite the fact that their purchases of that particular item might constitute a minor portion of their overall spending.

    Hafiz Noor Shams has a great post on how to calculate your personal inflation rate.

    3. Actually this is incorrect - wages have increased more or less in line with the CPI (about 2%-3%), not 2% above it, i.e. you're looking at nominal not real wage increases. And that's bad because it doesn't take into account increases in productivity which is about the same. On that basis, the average worker is in fact actually losing purchasing power over time relative to what they produce (i.e. labour is not being paid its marginal product).

    One interesting thing to note here is that this hasn't been a continuous process. The keys are the recessions of 1997-98, 2001 and 2009 where wages froze, but productivity and inflation did not. Those periods magnify even minor price increases, as they occur relative to stagnant nominal income. Even as wage increases began again after economic recovery, the implied one-time loss in purchasing power from those recessions have remained.

    ReplyDelete
  25. Sorry, the link to Hafiz's post wasn't working:

    http://maddruid.com/?p=9633

    ReplyDelete
  26. Yup I totally misread thestar report on real wages.

    Thanks for the comments and CPI links.
    Scanning through the table, it does seem to fairly tabulated. Food price is noticeably outpacing CPI. It will affect the urban low income more as food forms a bigger portion of their expense and they eat out more often(both parents working).

    ReplyDelete
  27. And the beauty of Pemandu's High income Economy is at least 7 mil will be earning less than 1.5k in 2020..that almost 50% condemned to poverty.

    Can we support ETP?

    ReplyDelete
  28. Where does this 7 million figure come from? It's not in the ETP.

    ReplyDelete
  29. Chk out Chapter 2 Pg 88.Projected wages 2020.
    As you well know Pemandu computes GNI as sum of wages n ebitda.

    ReplyDelete
  30. Thanks.

    You do realise that after adjusting for inflation, those numbers equate to a 7% drop in the ratio low income earners ratio to total employment over the next decade? I don't think that's bad at all.

    Of course the key question is, how much of that will be due to the ETP as against a base case of the economy left to its own. We'll probably never know that answer. Personally, I think those numbers might actually be met, but we might equally have gotten there without the ETP.

    I'm more interested in the ramifications on income inequality here, as while the ETP might take care of the labour demand side of the equation, we're still lagging on the supply side in building human capital. If those two continue to stay out of whack, we're going to see a continuing deterioration in Gini numbers to go along with the increase in GNI per capita.

    I'm less concerned over the poverty argument, as I think the loss of Ringgit purchasing power through inflation will be offset by currency appreciation. But I do think the battlegrounds shifting (ever so slowly) from absolute poverty to relative poverty i.e. inequality.

    ReplyDelete
  31. The bigger issue is the deception Idris Jala is foisting on us i.e "Let us all be rich. ." he says.Thats a huge lie cos even in ETP program 8 mil workers i.e 50% of workers will be poor.
    High income economy as per ETP rewards the rich.U chk out the wages of the 3.3 mil jobs created..its only 20% of created GNI.Thats not high income. .more of high profit.
    Am I flogging a dead horse?

    ReplyDelete
  32. No, you're not. I don't think he's lying, but I do think he's seriously deluded.

    On the other hand, 20% is an improvement compared to the current status - right now, it's just 7% (RM1500 and below) and 13% (RM2000 and below).

    ReplyDelete
  33. On the other hand, 20% is an improvement compared to the current status - right now, it's just 7% (RM1500 and below) and 13% (RM2000 and below).>>how do u work this out?
    My own calcs..total wages 2009 is between 190 bil-240 bil compared to the GNI of 670 bil>> 28% to 35%.High Income Econs eg Japan,S'pore etc hv Wages/GNI above 45%

    ReplyDelete
  34. Sorry, should have stated that those figures are shares of total household income. Which means as a percentage of GNI, they're even worse.

    Where I got the figures, I'm sorry I can't disclose.

    ReplyDelete
  35. Hi Hisham,

    Good that you pointed this out. Competency is critical, no matter who you are.

    Be great to hear your technical views on the speech's Big Claims - such as its critic on the economic growth rates projected by the Government, the number of those who would be poor and the rise in wages.

    Happy New Year.

    ReplyDelete
  36. Happy New Year to you Greg.

    As you know, 10 years out, it's not easy to evaluate claims objectively one way or another. Is it possible that wage rates between RM1500-RM2000 to be considered poor? I don't know...it's possible especially in an urban setting. Ditto with wage growth.

    Growth rates? My considered opinion is that we can get to high income status at below 5% real growth. Whether getting there actually means anything to the average man on the street is of course another thing entirely.

    ReplyDelete
  37. DS Anwar is not an Economy Graduate. No wonder if he say about inflationrate in malaysia , all number and percent always comes high.

    ReplyDelete