Singaporeans hurled abuse at the whole idea.

Lee justified his notion of appropriate salaries with reference to the benchmarking model established in 1994. But however they are calculated, to the public, the key issue is that the outcome is one of obscene levels of remuneration. There must be something wrong with the model. No use referencing it when it produces such results.

What is wrong with the model? Let me try to pick it apart.

But first, I had to force myself -- it made me sick to my stomach -- to read the 4 broadsheet pages of the Straits Times, 23 March 2007, devoted to apologia for the salary increase.

Having (in)digested that, this critique will comes in 4 parts, discussing various points that came to my mind.

• The conflation of civil service and ministerial salaries
• The MR4 benchmark
• The SR9 benchmark
• Increase in benchmarks since 2000

It strikes me as unproven why Singapore has to have a salary scale where political officers are lumped together with career civil service officers. In most other countries, they are on quite separate scales, enacted separately.

Source: Straits Times, 23 March 2007 Red arrows indicate benchmarked grades.

There is a case to be made for civil service salaries to be competitive with the private sector, after all a career is a career, but political positions are supposed to be filled by people who seek power out of passion and mission. It's not reducible to a mere career path. Arguments about ministerial salaries needing to be competitive with the private sector ring false.

Singapore seems to see government as a mandarinate, with ministerial posts as merely higher levels of the bureaucracy. By this schema, pay scales for ministers must top the bureaucrats' for dignity's sake, if nothing else.

We need to question the fundamental assumption that ministers must be paid more than their permanent secretaries or department heads. In the private sector, the CEO does not have to be the highest paid officer in the company. Many founder bosses work for less than their key lieutenants. 

The consolidated salary scale hinges on two benchmarks. In order to critique this problem, we need, painful though it may be, to understand what these benchmarks mean.

The higher benchmark is called MR4. It establishes the salary for ministers and senior permanent secretaries. Higher ups, e.g. Deputy Prime Ministers, etc, get more than the MR4 benchmark. [1]

To obtain MR4 for any given year, the government refers to the annual incomes of the top 8 persons in 6 fields (law, accounting, banking, multinational corporations, local manufacturing and engineering). These 48 individuals' incomes are then laid out on a linear scale and the median is computed. The median would be the midpoint between the income of the 24th and 25th persons in the range.

The MR4 benchmark is two-thirds of the median.

A bit of digression here. Others have pointed out that by this point, the model already introduces problems. The top 48 persons are rarely the same people year after year, but the government ministers and permanent secretaries do not change very much. This means that aggregated over a number of years, a particular private sector individual earns less than what it appears (since he doesn't regularly appear among the top 48). In other words, the model overstates the income of top professionals.

My contribution to the critique comes from looking at the numbers of persons involved. See diagram at right.

Consider this: There are 24 private sector individuals whose income, by definition, would be higher than the 25th person used to obtain MR4. But the Singapore government has 15 ministries. This means about 15 ministers and another 15 permanent secretaries. Not to mention Supreme Court justices and a few others whose salaries are also pegged to MR4 or higher. That makes perhaps 40 people. That means there are more people in the public sector with that kind of income than in the private sector.

Why should that be so? Why should the public sector be responsible for a disproportionate number of top-paid people in Singapore? And this is quite aside from the question of whether there is any sense pegging politicians' pay to career professionals'.

At this point we would recall the argument that public sector salaries must "keep pace", i.e. follow private sector pay. But is an employer who is the major employer in a given category the price-follower or price-setter?

The lesser of the two benchmarks is called SR9. The Straits Times (23 March 2007, page H14) said,

The entry-level superscale grade is SR9 and there is a salary benchmark for those on this grade.

Further on, on page H15, the paper said,

Good performers in the Administrative Service can reach that grade in their early- or mid-30's when they could be directors in ministries.

How is SR9 benchmarked? The government takes a look at the top 15 income-earners aged exactly 32. SR9 is pegged to the 15th person.

I wondered: Are there more than 15 SR9-level officers aged 32 in government? To estimate, I took a look at the organisational chart of the Education Ministry and the Trade and Industry Ministry. In the former, there were 17 civil servants of Director's level and above (but below Permanent Secretary). In the latter there were 10. Let's say the average is 13 per ministry. Multiplied by 15 ministries, there should be 195 civil servants of SR9 and above (but below Permanent Secretary).

This looks consistent with a brief mention by the Straits Times (page H13) that

There are now 230 AO's. [Administrative Officers]. In the past year, 7 retired, 13 resigned and one was transferred.

Of course, not all the AO's are 32 years old. They could be anything from 32 to 60, spread over some 28 years. Since there are 230 AO's in total, there should be about 8 AO's in each age cohort. [3]

Taking the age-32 cohort alone, we will see something like this pattern, where there are 15 individuals in the private sector earning SR9 or more, and perhaps 8 in the public sector earning SR9 or more.

The ratio's not so bad compared to the MR4 level. However, whether it is excessive for the public service to be responsible for 1 in 3 top earners aged 32 can still be debated.

Lee spoke about retaining talent. Salaries must be competitive, he said, and adjustments must be made when necessary. However, buried in the Straits Times' story (23 March 2007, Top govt salaries far behind private sector's) is this:

Today, [Administrative Officers in the 30s'] salaries are in line with the private-sector benchmark created for this group. But with salaries for young professionals moving up, especially in the financial sector, Mr Lee expects salaries in this group to 'soon come under pressure'.

Perhaps so, but with a recession more than possible in the US in the months ahead, don't take things for granted. In fact, the 7-year graph for the SR9 benchmark is sobering.

Annual income of 15th person aged 32

In the year 2000, the SR9 benchmark was $363,000. In 2007, it was $361,000, virtually unchanged. At the present moment, the pay scale for SR9 is $372,000, i.e. 3% higher than the benchmark.

Income of the 24th/25th person and the MR4 benchmark

So what was Lee talking about, when he said salaries were lagging behind? He was referring only to the benchmark figures for MR4, as seen in the graph at right. It has climbed to S$2.2 million when ministers' pay is still $1.2 million.

The question we should be asking is why is SR9 unchanged while MR4 has gone so far up? What kind of society is this where even those 3 or 4 rungs from the top (the SR9 group in the public and private sectors) have stagnating incomes, while those at the summit of summits fly off the chart?

* * * * *

I hope I have been able to show you the weaknesses of the benchmarking model. It is not a convincing basis for pegging salary levels. At the very least, the model has to be greatly overhauled. Furthermore, there is a case for benchmarking on a trans-national basis.

It is very telling how in almost all other matters, the Singapore government is constantly referring to international benchmarks to crow its successes, and by extension, its moral superiority, but never in this matter of ministers' and senior civil servants' remuneration.

Shouldn't Singaporeans know that in Britain, the salary range for Permanent Secretaries is £130,350 - £264,250 (S$388K – S$787K) [4], while here the prime minister thinks it ought to be $2.2 million?

I couldn't find on the web the pay range for Permanent Secretaries in Australia (where they are simply known simply as Secretaries), but I found that Senior Executive Service Band 3, which is one step below that of Secretary comes with a salary of A$190,402 to A$244,000 (S$233K - S$299K) [5]. Hence, we can roughly guess that the equivalent of a Permanent Secretary down under might earn something like S$300K - S$400K.

One can assume that the rest of the pay grades in the UK and Australian governments are scaled accordingly.

Does anybody think that the UK and Australian civil services are far inferior to Singapore's? How is it that those countries can have an effective public service with negligible levels of corruption at less than half the cost we have to pay?

Isn't getting value for money the responsibility of ministers?

© Yawning Bread