Friday, July 5, 2013

People First?

I read this article last week, but something about it really bothered me. It was like an itch I couldn’t scratch. I only figured out what was wrong yesterday (excerpt; emphasis added):

Time govt lived up to its slogans, says MIER chief

KUALA LUMPUR (June 28): Malaysia's economic policies need to be implemented in line with the government's slogan of 'People First,' which means the people should be the ones benefitting "firstly and mostly," the chief of the Malaysian Institute of Economic Research (MIER) said.

Unfortunately, this is not the case, pointed out Dr Zakariah Abdul Rashid, who is executive director of the think tank.

In an exclusive interview with, Zakariah touched on various issues, including the importance of interpreting economic indicators carefully, the widening gap between the rich and poor, doing away with race-based assistance and the importance of being competitive in choosing our leaders...

..."The economic indicators have to be interpreted carefully. An example is that when we recorded rapid growth in relation to other countries, our performance was commendable.

"The question was who actually benefited from the growth? There are two things, when we record GNI (Gross National Income) and GDP (Gross Domestic Product) per capita, this is a record for the whole economy. On whether it has trickled down to the people or not, this will be measured by the household income.

"Last year the real GDP has increased by 5.6% but household income in real terms only by 5.3% (for the period of 2009-2012). That means household income increased at a rate slower than the real GDP. That implies the benefit accrued to the household is less than the whole economy. This means that someone else other than households are benefitting more. Actually, we cannot pinpoint who is enjoying more. The policies, therefore, are not in line with the slogan of 'People First,'" said Zakariah.

I believe inequality is an important issue to tackle. Trying to prove it this way isn’t credible however.

There are at least two three four five pesky issues I can think of in this comparison of growth:

  1. The price deflators used to adjust household income and GDP growth for inflation are methodologically different. In other words, they measure very different concepts of inflation. For a basis of comparison, you should ideally use either the same deflator or forget inflation adjustment and use nominal numbers instead (either method would yield the same ratios). Without these tweaks, this becomes an apples to oranges comparison.
  2. Even then, you also still have to account for changes in household composition. The census data over the past three decades has shown household sizes consistently dropping, and this pace has accelerated in the period between the last two census’ (2000 and 2010). Extrapolating for the period 2009-2012, that means you need to make an upward adjustment to income of around 0.7% per annum to account for this difference. Otherwise, again, it’s an apples to oranges comparison.
  3. Similarly, you also have to take into account growth in the number of households, which again would need an upward adjustment to household incomes for a basis of comparison. The compound annual growth rate (CAGR) in the number of households between 2000 and 2010 was 2.9%.
  4. 2009, as someone likes to point out, was a recession year. Since in Malaysia unemployment barely budged but nominal GDP fell 7.4% there’s a fairly good case to be made that corporate incomes were depressed. Slower household income growth relative to other sectors during the recovery phase could then be interpreted as a return to normalisation, not a loss of income share. Income numbers from disaggregated national accounts data from DOS (unfortunately only available from 1999-2008) tend to support this interpretation, as household income tends to rise as a portion of national income during recessions and fall during a recovery. One way to get around this is to compare growth from a non-recession year, or from another year at the same stage of the business cycle.
  5. Since we’re looking at the distribution of income to Malaysian sectors, the correct comparison metric should be GNI not GDP.

To test my intuition, I checked the data. Nominal household income and GNI CAGR from 2004-2012 checked in at 5.54% and 5.56%. For the shorter period 2007-2012, the relevant numbers are 6.29% and 7.18%, for 2009-2012 7.50% and 9.71%.

Adjusting for household size and growth in the number of households dramatically change the numbers however, registering 9.37% (2004-2012), 10.15% (2007-2012) and 11.40% (2009-2012). That paints a very different picture of the situation, as it appears that household income is gaining relative to overall national income.

I’d take those adjustments with a pinch of salt however, as they are at best using rough estimates of the actual underlying changes in the number of households and household composition. Using the disaggregated national income data (which neatly avoids these complications) suggests quite a bit of variance in year to year ratios, but relative stability in the portion of income going to households.

What we don’t and can’t know from these growth comparisons is the distribution of income between households, which is probably a lot more important.

In short, rather ironically, “…the economic indicators have to be interpreted carefully”.


  1. Why not use the BRIM figures.
    Its solid figures shows that 7 million apply for BRIM that is below 3000 ringgit per month.

    Dont have to second guess anything.

    The numbers are there.

    1. Sorry, don't see the relevance of what you're saying. What has BR1M got to do with income growth?

      If you're thinking in terms of the distribution of income, that's available from the household income an expenditure surveys already. BR1M would in any case not be a very good guide to this, as retirees and pensioners (i.e. people outside the labour force) are eligible, even if they already have considerable assets.