Wednesday, September 21, 2011

Income Inequality in Malaysia

This is more of a first stab at the subject, as I haven’t had time to really play with the data yet. So there won’t be much in the way of analysis beyond eyeballing the charts. The source of the data is from the Economic Planning Unit, and appears to be aggregated from various Household Income Surveys (HIS).

Here’s the time series for Malaysia’s Gini Coefficient:


There’s gaps in the data because the HIS is only conducted once every 2-3 years – about twice within each development plan period. Looking at the observable pattern there was an obvious increase in income inequality through the 1970s, before dropping off sharply in the 1980s.

While I daresay some might equate this to the impact of the “infamous” New Economic Policy, and Tun Mahathir’s subsequent adoption of industrialisation in the early 1980s – I wouldn’t necessarily disagree without digging further into the matter – I’m also inclined to ascribe some blame to inflation. CPI based inflation peaked at over 17% in 1974 before dropping to an average of 4.5% in the second half of the decade. Inflation in the 1980s was lower still at an average of 3.2% (and nearly 0% between 1985-87). At this juncture though, I’m not confident enough to state definitively one way or the other, or what other factors might have been at play.

But it’s also interesting that income inequality has been essentially bouncing between .44 and .46 since 1987.

Since it’s a Malaysian tradition to break things down in terms of ethnic background, here’s the Gini for each community across the same sample period (click on the pic for a larger version):


More or less the same patterns as for the country as a whole, and the variation isn’t so much that I couldn’t put it down to sampling error alone. Malaysia’s Gini coefficient is generally higher than for any of the major communities, mainly due to extremely high inequality within the smaller communities. This is understandable – we’re talking here about everyone else from foreign professionals to Bangladeshi kitchen workers.

There’s also the question of income disparities between communities, for which data is also available:


No big surprises here – for the most part the average Chinese household has earned more than the average Malaysian household, Bumi households earn less, while Indian households tend to be right on the average. Other communities are all over the place, possibly reflecting the changing composition of foreign immigrant labour.

But the ratios are what most people talk about so I imagine this next chart would be of interest to many:


As of 2009, the average income of Bumi households is within 10% of the Malaysian average, and within 30% of the average of Chinese households. Most of the gains were achieved in the period before 1987, but there’s also been considerable closing of the gap in the last decade. I haven’t yet looked at the data by income stratum (which is actually available), so generalising from this might be premature.

One last chart to show – I matched the income data from EPU with household expenditure data from DOS (released just last month). The expenditure data only covers from the 1993/94 HIS onwards, and there’s only four data points to look at. Again difficult to generalise, but it’s an interesting chart nevertheless:


Mind you, this doesn’t say the cost of living has fallen, just that nominal income growth has exceeded nominal expenditure growth. And the experience by income stratum might well be different, though unfortunately DOS wasn’t kind enough to release that depth of data.

Note that what we’re looking at here is primarily household income and expenditure data. There’s no data on wealth inequality that I know of, which might shed more light on the actual state of things in Malaysia.

Technical Notes:

Household Income and Poverty, Economic Planning Unit, Prime Minister’s Departmentc (accessed 21/9/2011)

Household Expenditure Trend 1993/94-2009/2010 from the Department of Statistics (warning: pdf link) (accessed 21/9/2011)


  1. "And the experience by income stratum might well be different, though unfortunately DOS wasn’t kind enough to release that depth of data."

    bro hishamh

    Have a feeling its really bad at the lower end of the curve.....this is believe is one of the major issues currently facing our nation. And if they split rural vs urban lagi evil i think

    When i was back for the holidays I saw ferarries, maseraties everywhere...I asked myself..zaman gua kecik2 dulu nak tengok sebijik maserati pun susah..paling kuat tengok gambor maserati boomerang dalam playing cards.....

    I feel that the Gini numbers are underestimated if you count the "hidden values" of underreported incomes, tax evasion, Sdn Bhd coverline beli assets n kereta for individuals etc...

  2. Tak paham..

    If wages is only 30% of GNI the balance 70% are operating surplus(EBITDAs in Pemandu-speak)
    OS is shared by small group of folks i.e the business guys,asset rich guys etc,,which I reckon is not more than 20% of population.So 80% is sharing 30% i,e abt 200 bil by 22 mil & 450 bil by 6 mil folks.
    So its 9k/yr versus 75k/year per capita.GINI is in the high 50s.
    If you look at the wage structure (ETP guidebook Chapter 2) and do some back of envelope calcs u will be able to reaffirm these numbers.
    I tell you we hv a huge disparity.Its too sensitive to discuss.But the mentality of our leaders will further increase the divide,Paying GLC CEO 100 times entry level grads (running a monopoly)is pure senseless greed.

    The only reason why the gap is not so evident is the easy credit (unlike Egypt).But if things continue to go up n salary frozen at minus 2% real growth..the meltdown point will be in 2015.And the meltdown will not be dissimilar to Egypt Workers revolution..Fighting for 200% salary adjustment just for survival.

    So,pls..we need major restructuring..wages must be 45% of GNI minimum.W/out that its a failure mode.And leaders/CEO must curb their greed.Keep CEO/Entry level to below 30 pls.

    I hope I am making sense

  3. @satD,

    We'll call it the supercar index!

    But seriously, I think that's more a factor of rising incomes in general. I'd take underreporting of income as a given especially at the top end (it's a global phenomenon), but adjustments that I've read of don't move the numbers very much. So I don't think that invalidates the trend in the data - essentially inequality has been static for the last two plus decades.

    I've just looked at the share of total income by income level, and there's been little overall change since (again) 1987. Before that gains were made by the urban middle 40% and the overall bottom 40%, at the expense of the top 20% generally, but particularly from the rural rich.

    Having said that, the top 20% income earning households still grab 50% of the pie.

    I'd be really interested in looking at the earnings of the top 10% or even 1% - that may be a better sign of things.

  4. @anon,

    You're probably correct in your calculations, but I think the discrepancy arises from conflating wages with income in general. The data on inequality in this post is based on household income surveys, which will have both wages and capital components. The GNI (I'd have to check the handbook to be sure), uses calculates employee compensation i.e. wages alone.

    Why I think this distinction is important is that I think a significant proportion of the lower income households will have income classified wholly as capital income, not wages. Income from enterprises, sole proprietors, and partnerships would count as capital income. That includes a huge span of different activities, from PriceWaterhouseCoopers to your corner grocery shop to your roadside mechanic.

    As of June there's over 4.5 million registered businesses in Malaysia (compared to under a million registered companies) - if even half of those are active, we're still talking a significant portion of the population whose earnings would be classified as capital income (even if it isn't very high), and won't be included under the wages calculation under the national accounts.

  5. Hisham,

    Would you happen to have median data as opposed to mean data? Would give a better sense to where the middle class really lies as opposed to mean incomes being skewed by the obscenely high incomes of the rich.

    I am interested to know the distribution of income amongst the population (e.g. top 20% get x amount, next 40% get y amount etc.) in general and by ethnicity.

    Also some time series data would help us see whether the rich are getting richer.

    But overall, interesting post, and I think this ties in well with you other posts on achieving a high-income (and high-cost) nation

  6. Rodger,

    You can hit the EPU link at the end of the blog post for more. They don't provide median numbers but the percentile breakdowns for top 20%, middle 40% and bottom 40% are available. Or you can refer to my reply to satD just now.

    I wish I could peek at the raw data - as it stands, you don't really get much of a sense of how things really stand, except that it looks ok (!?). Malaysia's Gini is and has been around the global median.

  7. You can check with DOS guys working on the GNI numbers..this income category is abt 10% of GNI and as you correctly state is not in the wages category.

  8. Mixed income not ez number to derive but best estimates is 10%.Add that wages stratum n corresponding 2 mil recipients,the Gini will be even worse off.The supercar index is real sign of how rich the super rich are.And the burgeoning hsehold debt indicative of how poor the rest are.By 2020 with jalanomics,the mid income will eescape the TRAP...n go low income.