Tuesday, April 2, 2013

Torturing The Data Until It Confesses

I don’t get it. Why do people think that nominal figures at any given point in time mean anything? Actually I do get it – it’s called money illusion. And perhaps a dollop of the concept of “natural” prices (though that’s another thought for another day).

Case in point (excerpt; emphasis added):

Why Is Genius PM Najib Delaying 13th General Election?

…Since he’s toast any way you look at it, why not spend the remaining precious time maximizing contracts allocation to his cronies? Well, actually Najib is a genius primarily because he managed to increase the country’s debt to RM502 billion last year, the highest in history and reaching 54% of GDP. In short Najib doubled the debt in 4-year what Badawi managed to do in 8-year, an impressive milestone indeed. If that is not a sign of genius, I don’t know what is (*grin*). It’s critical to build his war chest, just in case he miraculously gets to keep his premiership. If somehow he managed that, his next battle in UMNO election needs billions of dollars to secure the presidency seat…

The graphic accompanying the piece shows the increase in government debt since 1990 – don’t bother checking, the numbers are all correct. The interpretation, however, is not.

Quite aside from the issue of the macro-stability function of government spending, using absolute nominal values as a measure isn’t terribly meaningful in any economic sense (though I’m sure some would say it makes political sense). Saying debt is at a historical high in absolute terms doesn’t convey any relevant information at all.

For one thing, it seems incredibly hard for people to grasp that RM1 in 1990 and RM1 in 2010 have very different purchasing power due to inflation, and aren’t directly comparable.

For another, many economic variables hit “historical highs” every year, which isn’t exactly indicative of whether you’re doing well or not. Otherwise, it would be quite appropriate to claim that since GDP hits a new historical high almost every single year, the government must be doing a good job:


The economy now is 8 times larger than it was in 1990, while government debt has only increased less than 6 times. But would anybody be prepared to claim that this shows good governance and good economic management? I didn’t think so, and they would be quite right. The business of government and managing the economy is a lot more complex than that, and subject to many varying forces, most of which are outside policy-makers control.

I could point out that the debt to GDP level was 79.5% in 1990 while currently its only 54%. Does that make Anwar a bad finance minister and Najib a good one? No again.

Just comparing absolute values across time is neither useful nor appropriate. Even relative and/or adjusted values need some care. Just making a comparison from one variable of one agent in the economy isn’t sufficient to make a judgement of the appropriateness of its current level, or dare I say it, the competency of its managers. Government borrowing and debt level are wholly intertwined with demand, supply and monetary conditions in the rest of the economy. Context matters.


  1. Haven't had a chance to post a meaningful comment in a while.

    If we want to go on actual scorecards we can say that the administration (not blaming the Finance Ministry or any specific individual here...in my mind the burden or credit is shared equally) in the early 1990s set the stage for the '98 crisis. But really, it's the same administration that orchestrated the rebound.

    But it is interesting to note that public sector debt (absolute values) stagnated/declined during the early 90s in line with the marginal (and only) fiscal surpluses Malaysia had. Combine that with a current account deficit, one might say that the private sector over-leveraged during that decade and that it really isn't viable for Malaysia to return to those GFCF (%of GDP) levels we once had, despite what policymakers or their detractors say....but then again, that's my 2 cents.

    But I agree with you, there needs to be a better metric to judge a finance minister or the administration (or other key policymakers)

    1. Jason

      I dont like being partisan and neither is Najib exactly my cup of tea excepting his economic reforms/initiatives bar his tinkering with the NEP. His stance on national Security (ISA repeal etc) troubles me but despite that he will still get my vote. But let us be impartial about certain issues:

      a. "If we want to go on actual scorecards we can say that the administration (not blaming the Finance Ministry or any specific individual here...in my mind the burden or credit is shared equally) in the early 1990s set the stage for the '98 crisis. But really, it's the same administration that orchestrated the rebound."

      Wrong: Not exactly the same admin pre and post 1998 with radically different economic ideologies to boot. Save to say the stage for 1998 was already presaged with the failed cap controls on inflows in 1994 and 1998 was exacerbated by one man calling the shots for about three quarters or so as Jomo points out in here (bear in mind Jomo has certain a bias too (wink):

      Pathways through Financial Crisis: Malaysia (available in pdf format)

      2, Public sector debt stagnated or declined because public sector expenditure was curtailed and the savings wre used to offset the debt. But what that will not tell you was the impact of diminishing public expenditure in areas such as agriculture and rural development and the subsequent attendant impact on icomes and consequently on the Gini. In fact, 1997 recorded the highest Gini peak what with crimped rural incomes contending with burgeoning urban incomes driven by the facilitation of speculative hot inflows in equities and property. As to why, go figure the prevailing ethnic dynamics then.

      Hisham, you hit the button on something I have noticed but I have a slightly diffrent take which i will post about morrow......

      Warrior 231

    2. Warrior,

      You're rather fond of using the word "wrong" aren't you?

      But I agree that there was an about turn in macro policy coinciding with the pre-Anwar/post-Anwar era, though as you say, the man in charge stayed the same. In that sense, Jason isn't "wrong".

      Two other things: Your link to the Jomo paper doesn't work; and I'd like to see some backing for your assertion that the Gini peaked in 1997.

      Other than that, I'd be interested in hearing your thoughts

  2. Yeah, I was about to agree with Warrior (about the Gini) until Hisham said he was suspicious.


    It says Gini peaked in 1976. Warrior, do you have another source? I wouldn't mind checking it out as well.

  3. 1. Wrong? What is so wrong with “ wrong” mate? You seem to flinch at its leprous sight...chill out dude. Wrong is the ultimate kick a&%e to smug complacency. I grew up with” wrong: hollered down my ear canal, banged together my tympanum, jangled my anvils, burned my cochlea before singing my synapses. I learned to avoid it like plague but even when it sneaked up and pounced on me unawares, I learnt to appreciate its seductive call as it yanked me away from the torpor of many a comfort zone and drove me, like a slave driver would his slave, to search for Mr Truth. That is a warrior credo: subservience to Truth (only birthed by objectivity and commitment to empiricism.)

    In any case, it jives in with my passionate love for all things autocratic, slavegirls included hahahahahaha (democracy is an inefficient use of time and resources, no doubts about it!). For you see, like all true-blue autocrats, something is either right or wrong and that by itself is a life simplifying process as much as it is a cleansing act of fallacious thought-webs. So dude, embrace wrong, strip her of her faulty generalizations, titillate her falsehood, make love with her fallacy so that you attain the orgasmic delight of truth. And moan in ecstasy as paroxysms of reality washes away the dusts of illusion...hahahaha...It will set you free, man but not as efficaciously as a quality weed would! (ROFL)

    2. Ok Jason of the Argonauts, I admit I was WRONG. Need not trot out “suspicious” to feather and tar a simple uncomplicated gentleman in cybersquare for after all I don’t deliberately assault Mr Truth and a wrong is a wrong and I accept it, period..full stop..enough said, cukup..noktah hahahahaha.

    I was relying on this source:


    and the only caveat I will use with regard to ur 1976 figure is that the GINI then was the consequence of an externally induced shock event (remember : the Oil Shock of 1974, OPEC blah2) and was registered merely 6 years after the NEP which implementation by the way had stalled somewhat with Razak’s ailment and Hussein Onn’s (he wasn’t agog of the NEP by the way) ascent. But by invoking the caveat, I would be going back on my ‘acceptance of wrong”, wouldnt I? trying to weasel outta a tight corner with your headlights beamed in mine eye, am I?(ROFL)? , So let the thar chilly wind blow for its no skin off the Warrior’s nose.............hahahahahaha.

    But the centrality of my contention, still stands, right? = 1997 Gini is a by-product of deliberate policy failure and wilful neglect of a certain ethnic (let me "race soup" it up a bit, for after all, me like Paolo di Canio of Sunderland, are racists right, errr, bad influences on society......hahaahahaaha).

    Just to ram home my point about comparative performances, look at the education Gini:


    So much for someone and his pals peddling snake oil everywhere.

    3. Back to Hisham...still owe you my thoughts on the crux of this thread. Will put it up soonest, rest assured.

    Warrior 231

    1. ROFLMAO!

      All that for "oops, I made a mistake"? You could've just said "sorry, my bad", it would have been about 500 words shorter!

      But thanks for the links. Good stuff.

      BTW, there isn't enough variation in the Gini numbers to establish a link with policy direction, but then given how the Gini index is actually constructed it wouldn't be a good marker for your hypothesis either. The Gini measures within group inequality, not between group inequality, which is what you're aiming at. You're better off just sticking with the actual household income numbers.

  4. oops ...forgot Jomo the useless Gomo:


    Warrior 231