A bird’s-eye view of the budget – Addendum

This post is in response to the first few comments received for the post A bird’s-eye view of the budget. If you haven’t read that post, this one here may not make any sense to you; you should read it first before coming here.

I’m rather pleased with the comments I’m getting, as they indicate that readers are looking at the numbers. I think this is to be encouraged. A habit of referring to facts and figures makes for more grounded discussion than just spewing criticism of the government (as some websites tend to encourage).

So let me now take the issue further by responding to a few of the more cogent comments.

Defennder said in his comment to the earlier post:

Note that GST revenue did not decline because of the recession either. So there must be more than meets the eye than just pointing out that personal tax revenue did not decline meant that the rich had insulated themselves from recession. Income tax revenue is projected to decline in 2010, although I have no idea why.

Actually, GST collection did decline. It is not obvious if one looks only at the gross amount of GST dollars collected, for this would show the quantum increasing every year from 2006 to 2011. However, as I mentioned in the earlier article, the GST rate changed significantly in 2007. Prior to 1 July that year, it was pegged at 5 percent. After that it went up to 7 percent.

Hence, it is necessary to pro-rate all the figures to the same GST rate in order to compare if consumption — which is what GST reflects — declined in any year.

For Fiscal Year 2006, we can do a simple adjustment from 5 to 7 percent. As you can see from the table below, if in 2006, the GST rate had been 7 percent, we would have collected S$5.57 billion instead of S$3.98 billion.

Adjusting FY 2007’s figure is a bit more complicated. The fiscal year starts from 1 April. The rate hike took place three months into that fiscal year, on 1 July 2007.  The average GST rate for FY 2007 is therefore 6.5 percent. Doing a similar pro-ration, we would have collected S$6.64 billion in GST if it had been 7 percent throughout that year instead of the S$6.17 billion actually collected.

In FY 2008, when the rate was 7 percent throughout, we collected on S$6.49 billion, which was 2.3 percent lower than what would have been the comparable figure for FY 2007. That’s your recession.

* * * * *

As I pointed out in the earlier article, one can see the recessionary dip in all the broad-based sources of revenue — Corporate Income Tax, Assets Tax, Customs and Excise Tax, Dividends and Interest from investments, and GST. But there is one exception: Personal Income Tax (inclusive of Withholding Tax). This combined line item kept rising year on year. It really begs explanation.

I had suggested that it indicates how insulated the richer segment of our society is from the vicissitudes of economic cycles. They are able to maintain their personal income despite the recession. Is this a measure of their ability to manipulate income and wealth distribution to their own advantage, I asked?

A comment by suggestion said we should look at population growth, a point also made by John Tan. Suggestion wrote:

That’s gross income tax and not per capita or household tax?

The population grew a lot from 2006 to 2010

A quick look at the Statistics Department website reveals that the population midyear 2006 was 4.40 million. The total population midyear 2010 was 5.08 million. That’s an increase of 15.5 percent.  Given the one year lag in income tax assessment, we should look at Personal Income Tax (+Withholding Tax) collected in 2007 and estimated for 2011. The figures are S$5.69 billion and S$8.16 billion respectively, an increase of 43 percent.

It thus seems that population growth alone (per capita income tax holding steady) cannot fully explain the increase in personal income tax collected; it would have to be population growth + significant increase in per capita income leading to higher per capita taxes payable  — and this while a recession ravaged Singapore in the interim. In any case, I think a huge part of the population increase was in the form of low-wage workers who do not pay tax.

But we know that a large segment of the Singaporean working population suffered wage restraint through the recession, so it suggests that a minority of our working population enjoyed really good years despite the recession, in order to create that rising average. This was what led me to speculate that the rich were able to insulate themselves from the economic cycle, maintaining their upward march without a hiccup.

Sprechen Sie Singlisch contributed an important perspective:

quite a bit of your analysis hinges on who exactly these middle to high income persons are with respects to the population in Singapore (30%, 50% or 70% of the population better yet a histogram). Any data with respects to the income distribution population would surely strengthen your argument.

Also, I suspect the brunt of the recession fell on transient workers who I believe make up the majority of low income workers here. This does not substantially change your argument “that pain is disproportionately borne by other sections of society” but it would change the moral calculus for many Singaporeans.

Ah! Histogram. If only it were so easy. The first problem is that data regarding incomes is hard to find. Our Statistics Department is very bad at providing comprehensive data consistently over the years. Although a census was conducted last year, I cannot find any report from there about earnings; maybe it’s coming.

All I could find was data from the 2000 census, i.e. ten years ago. Source.  The numbers are out of date, but since what we’re interested in is percentage distribution across different income bands, they are still usable in a way. Relative distribution would not have changed much over a decade even if absolute numbers might have changed.

However, while the curve might keep the same shape, it will have crept up the income bands over the past decade. On the other hand, since I am unable to make any adjustments for tax reliefs, they kind of cancel each other out; the chargeable incomes of 2010 is probably similar to the gross incomes of 2000. In the table (thumbnail at right) you will see the assumptions I made regarding average incomes and tax payable on them.

Based on those assumptions, the following graph shows how much of Personal Income Tax collected comes from the top end of our society:

I estimate that about 50 percent of the Total Personal Income Tax collected by the government comes from the top 3 or 4 percent of income earners. 80 percent of Total Personal Income Tax collected by the government comes from the top 14 percent of income earners.

Therefore, in response to Sprechen Sie Singlisch’s comment, the middle to high income group that largely determines how much Personal Income Tax flows into State coffers is probably about one-seventh of the working population. It is quite conceivable that salaries for two-thirds of the population can stagnate, while the top 15 – 20 percent see strong income growth even through a recession, thereby paying increasing income tax year on year.

The thumbnail on the right is the Personal Income Tax structure, just for the record.

13 Responses to “A bird’s-eye view of the budget – Addendum”


  1. 1 stngiam 21 February 2011 at 21:53

    Actually, MOM does publish income data in its Labour Force Report. It’s just that the data is hidden amidst useless stuff like training indicators. To their credit, they do provide the data in Excel format, see

    http://www.mom.gov.sg/statistics-publications/national-labour-market-information/publications/Pages/statistical-tables-2010.aspx#lf

    Unfortunately, the MOM data, just like the DOS data, only tells part of the story because they don’t include foreigners. So the top and the bottom of the wage distribution are cut off.

    Despite that limitation, the data does make it clear that making >$10K a month just isn’t that exclusive any more … From about 4% of the workforce in 2006 to 6.2% in 2010 (Table 30)

  2. 2 xtrocious 21 February 2011 at 22:39

    Regarding the part about the rich insulating themselves and getting richer…

    This is based on anecdotal evidence (n=1) but those people who made money in the stock market during the 2009 crash were the rich retail investors i.e. HNWIs who could afford to pick up prime blue chip stocks on the cheap while the rest of us were tightening our belts in preparation for the worst (which fortunately never came)…

    But even then, capital gains do not attract personal income tax…

    Hmm…so that’s one explanation out of the window…

    Of course, those who serviced these HWNIs – like brokers, housing agents, car salesmen etc would have made a lot of money as well – so that may account for part of the income tax increase…

  3. 3 Hong 22 February 2011 at 00:40

    anecdotal evidence that the rich are doing very well?

    Luxury cars a hit despite increase in COE premiums

    http://www.channelnewsasia.com/stories/singaporebusinessnews
    /view/1112147/1/.html

  4. 4 Tan Kuan Han 22 February 2011 at 11:13

    A foreigner earning the similar amount of salary as a Singaporean will have to pay more income tax as CPF employee contribution, 20% of the salary package is not taxable income.

    Hence even if we exclude other tax reliefs enjoyed by Singaporeans,

    the larger the percentage of foreigners(tax resident status) in our workforce, the higher the amount of total chargeable income in Singapore.

    For example, if a foreigner and local earn $30,000 assessable income per annum,
    excluding tax reliefs,

    the local will have $30000-20% CPF contribution = $24000,

    with a chargeable income of 4000, yielding $140 at the 3.50% tier rate.

    Whereas the foreigner will be paying $350 with a chargeable income of 10,000 dollars.

    I think the percentage of foreigners in the workforce has been on an upward trend prior to 2010, hence this may be one possible reason why total personal income tax revenue had gone up as a whole?

    Also, if we take your methodology of prorating the GST revenue figures to the prior GST rates in order to compare for consumption,

    shouldn’t we also prorate the income tax figures too to include income tax rebates as according to haveahacks’s example in your prior article?

    Reproduced here(sorry for not asking your permission first,sir/madam):

    “haveahacks
    21 February 2011 at 21:06
    Oops. Partially take back what I said earlier. According to IRAS’ website, there was a 20% tax rebate in YA2008 and 2009, plus this year. That means that the income tax take also should be adjusted to what the taxes would have been without the rebate.

    Tax paid in 2007 2008 2009 2010 2011
    on income earned in 2006 2007 2008 2009 2010
    $billions 5.7 8.2 9.1 8.1 8.8
    (Assuming no income tax rebate)

    In addition, income tax is paid on income earned in the previous year, so after making these adjustments, we can see that incomes did in fact fall quite sharply in 2009. It’s just that that was masked by the withdrawal of the 20% rebate in 2010. We can only speculate as to whether there would have been a rebate if the election had been due last year rather than this year…..”

  5. 5 UnimpressedBudget 22 February 2011 at 13:22

    Have you taken into account that FY08 was the year of financial meltdown and many mid/high level employees were retrenched across boards? As a result, there could be retrenchment and bonus payouts to these people (which I suspect is large) that contributed to the high collection rate in FY09.

  6. 6 Sprachen Sie Singlisch? 22 February 2011 at 14:59

    Hi Alex,
    thanks for pulling the distribution data and crunching the numbers. stngiam too for the MOM updates. The last bar on the right is certainly impressive! So yes I have to agree that income tax collected is dominated by the uber-rich top 15%. Now its just a matter of thinking through those tax relief/rebates, at least the major ones point out by haveahacks and others.

    “A quick look at the Statistics Department website reveals that the population midyear 2006 was 4.40 million. The total population midyear 2010 was 5.08 million. That’s an increase of 15.5 percent.”

    Like to point out that this number includes the many low income transient workers who don’t pay income tax. This 15.5% increase is an upper bound and the actually effects on income tax returns will certainly be lower. Also, notice how a lot of unknowns revolve around the non-resident population information?

    btw no matter whether the claim holds up, this has been a very enlightening discussion, not to mention the rich source of data that has been dredged up in the process that I was completely unaware of.

  7. 7 Fox 22 February 2011 at 17:40

    “The total population midyear 2010 was 5.08 million. That’s an increase of 15.5 percent. Given the one year lag in income tax assessment, we should look at Personal Income Tax (+Withholding Tax) collected in 2007 and estimated for 2011. The figures are S$5.69 billion and S$8.16 billion respectively, an increase of 43 percent.

    It thus seems that population growth alone (per capita income tax holding steady) cannot fully explain the increase in personal income tax collected;”

    It is also possible that the 15.5 percent increase in population is disproportionally made up of economically active foreign workers (few children and retirees). Hence, the increase in the working population is probably significantly greater than 15.5 percent.

    What you should look at is the labour force size. It was 2,594,000 in 2006 and 3,136,000 in 2010, amounting to a 21 percent increase.

    Also, it is rather unlikely that this increase is from mainly low income transient workers. The educational composition of the 2010 labour force (57.7 percent) has signiciantly more people with post-secondary education than the 2009 labour force (51.1 percent). This increase over just a single year is actually quite significant.

    Don’t forget that the economy grew very strongly in 2010.

  8. 8 Tan Kuan Han 22 February 2011 at 20:52

    Opps, I didn’t notice the CPF contribution cap of 4.5k/6k wage ceiling.

    Perhaps the effects of tax reliefs for Singaporeans plays a smaller part in affecting tax revenue than I previously thought, sir.

  9. 9 Jason 22 February 2011 at 22:31

    Budget 2009 estimated income + withholding tax collection for FY2009 to be $6.2bn. But by Budget 2010 revised income + withholding tax collection was up to $7.3bn. Thus, the jump was not expected by the government when Budget 2009 was released.

    Certainly unemployment was not as bad as forecast, and wages (and thus tax receipts) were being supported by the Job Credit scheme. Overall labor force and employment grew slightly. Those things go some way to explaining the jump in 2009.

    Note that the reverse seems to have happened in FY2010. Income + withholding tax were initially estimated to hit $9bn, but revised estimates are now $7.4bn. I am not sure what is the explanation for this.

  10. 10 defennder 23 February 2011 at 15:13

    Hi thanks for the clarification and analysis. I’m going to have to think harder about this before I get back to you.

  11. 11 David 23 February 2011 at 23:01

    @ Hong, 22 February 2011 at 00:40

    That demand for luxury cars increased despite increase in COE premiums isn’t really a good anecdotal evidence.

    The above-mentioned observation is the result of economic forces at work. In short, the demand for luxury cars by the rich is generally price inelastic since COE premiums constitute a small proportion of their income. Hence even with the increase in COE premiums, many of them don’t really feel the pinch. Whereas, the middle class is much more price sensitive, the increase can be a significant deterrent.

    For more information, you may refer to http://www.comp.nus.edu.sg/~ipng/oped/2011/20110203ST_COE.pdf.

  12. 12 T 25 February 2011 at 10:52

    /// A quick look at the Statistics Department website reveals that the population midyear 2006 was 4.40 million. The total population midyear 2010 was 5.08 million. That’s an increase of 15.5 percent. Given the one year lag in income tax assessment, we should look at Personal Income Tax (+Withholding Tax) collected in 2007 and estimated for 2011. The figures are S$5.69 billion and S$8.16 billion respectively, an increase of 43 percent.

    It thus seems that population growth alone (per capita income tax holding steady) cannot fully explain the increase in personal income tax collected; it would have to be population growth + significant increase in per capita income leading to higher per capita taxes payable — and this while a recession ravaged Singapore in the interim. In any case, I think a huge part of the population increase was in the form of low-wage workers who do not pay tax. ///

    Alex, I think it is the other way round. The increase in resident population is largely made up of new citizens and PRs – people that Singapore want – in other words, fairly well paid white collar professionals and other PMETs. And all these are effectively tax-paying and paying relatively higher taxes compared to the previous stock of citizens which include retirees, unemployed, children, and students. So, the % increase in tax collection should be much higher than the % increase in population.

    • 13 yawningbread 25 February 2011 at 11:58

      I’m sorry I didn’t make it clear. The figures I gave (2006: 4.40 million, 2010: 5.08 million) referred to total population, which includes work permit permits, foreign students, non-working wives, etc. It increased 15.5 percent as I said, over those 4 years.

      For Resident Population (i.e. citizens + permanent residents), the figures are: 2006: 3.53 million, 2010: 377 million representing an increase of 6.8 percent.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s





%d bloggers like this: