Monday, October 18, 2010

Budget 2011: By The Numbers (Updated)

I’m not going to regurgitate the analysis and feedback available through the papers since, well, it’s available through the papers (here’s the Star’s reporting).

But I do want to give some perspective to the government’s spending plans and forecasts for this year and next, which are of greater interest. Quite frankly, apart from the front-loaded news on the infrastructure projects, this year’s budget presentation is about as anticlimactic as they come.

But let’s see the government’s spending plans and revenue projections in historical perspective (RM millions; sample: 1976-2011f; includes development spending):


There’s been quite a bit of criticism over the 6.8% log increase in operating expenditure – I think it’s a little misplaced. Take away the last two years (2009-2010), and that’s the slowest increase since 1999. So even with the mildly populist bent of this budget, the PM hasn’t exactly thrown the pocketbook wide open (log annual changes; sample: 1976-2011; not including development expenditure):


Looking at it from another perspective, here’s the natural log-transformed series of both revenue and expenditure for the first chart above:


A linear (straight) slope implies constant growth in this chart, but both revenue and expenditure – including forecast and planned for 2011 – are way below trend over the last three years of the sample.

And just to rub it in, revenue and expenditure scaled to GDP (using government forecasts for nominal GDP for 2010-2011):

04_r_gdpOperating expenditure hit an all-time high as a ratio to GDP in 2009, but has fallen off to more “normal” levels in 2010-2011. Still high, but not so bad.

Of course, all this depends on whether the official growth forecasts are accurate. I’d say there’s a pretty good chance that the 7% target this year will be met, but I’m not so sanguine about next year.

I did a few calculations some time back on the path of the economy over time (see this post for the technical discussion and model particulars). It occurred to me that since we’re essentially off the recovery path and back to a “trending” economy (a reasonable assumption), the stochastic models I used could be employed as forecasting models (real GDP; sample 2000:1-2010:2):


So I’ve updated the model to include the 1H 2010 GDP numbers, reran the regression, and forecast quarterly real GDP out to 4Q 2011:

06_forecastAnd in table form:

  % Y-o-Y % Q-o-Q, SA, annualised
3Q 2010 5.7% 0.0%
4Q 2010 3.6% 3.1%
1Q 2011 4.3% 8.7%
2Q 2011 3.4% 1.6%
3Q 2011 3.8% 2.1%
4Q 2011 4.1% 4.0%

The full year point forecast for 2010 is 6.94%, a little under the government’s growth forecast – next year however is a disappointing 3.87% overall. I’d also draw attention to the flat sequential growth in 3Q 2010 and the less than stellar 4Q forecast.

I’d be the first to point out the uncertainty factor in these forecasts – a 95% confidence range yields full year growth estimates of 5.6%-8.3% (2010) and 2.3%-5.4% (2011). That’s a fairly wide range, and put’s the government’s forecast for next year right at the top of the model’s forecast range.

In other words, getting to 5.0%-6.0% growth isn’t outside the realms of probability, especially if I specified the model wrongly and there was a change in the slope and not just in the intercept. It’d be interesting to see how accurate this exercise turns out to be.

Technical Notes:

Government revenue and expenditure forecasts from Economic Report 2010-2011


Hafiz asks if the substantial increase in subsidies in the middle of the decade changes anything. It’s a good point and worth investigating. It’s hard to tell, since I haven’t got the detailed forecast breakdown of government expenses for this year, but there’s a strong possibility he’s right – as usual.

But here’s the relevant charts:

In RM millions (the “official” numbers, not Pemandu’s):

07_subsidies As a ratio to nominal GDP:

08_subsidies_r Comparing gross against net (RM millions):


09_levels And comparing growth rates:


The sharp spike in 2008, when crude oil hit over USD140 per barrel, is the main culprit. Net government expenditure rose further in 2009, but there’s not enough data for 2010-2011 to make a full determination of whether there has been a net increase in spending over the same period that total government expenditure has been slowing.

The level of subsidies are a little higher this year, but there will be some adjustment downwards given the small cuts July past. On that basis, and if there’s another unannounced cut in subsidies forthcoming in the coming year, it looks like spending is basically on trend.


  1. I think there's a need to isolate the effect of high fuel price and the associated cost of subsidy that has been phrased out. I have a feeling that that huge increase in the middle of the decade was due to high crude oil prices.

    I think a comparison of 'core spendings' (core spending that involves actual spending on government) would provide a better picture of government expenditure rather than full spending that included subsidy that is no longer there.

    I'd bet the growth rate hasn't changed much. I'd think that the mid-decade was full of extraordinary events that its effects should be controlled.

  2. Hafiz, I've updated the post with the relevant charts - looks like you're right.

  3. Overall, i'm worried about Government plan to consolidate its account to below 3% by 2015 (as outlined under 10th MP) given the still-high deficit-to-GDP number (@ 5.4% in 2011, only slight reduction from 5.6% this year). It seems like the plan need to be back-loaded.

  4. To be perfectly honest, I'm completely unconcerned about the deficit. As long as the debt to GDP ratio stays below 60%, I think we'll be just fine.

  5. Any inputs on new sources of economic growth?

  6. No, and I don't think that will be a feature of budgets going forward. That will all be under the NEM/ETP where the private sector is supposed to lead (with the govt acting as cheerleader).