Getting data from the Singapore government often feels like trying to draw blood from stone. Information comes out in dribs and drabs. At different times — and sometimes even in the same report — the numbers refer to varying and incongruent time periods. Another frustration is that sometimes numbers are disclosed, while elsewhere on the same page, it is percentages that are mentioned (without disclosing the base numbers). Terminology is left undefined and one suspects that the meaning of even the same word can vary from one report to another, perhaps depending on political expediency.

Read this paragraph 52 from Second Minister for Manpower Tan See Leng’s speech at the Committee of Supply debate on 3 March 2021:

Mr Pritam Singh asked for the details of the number of migrant workers who were underpaid, and details of restitution made. Between 2015 and 2019, 950 errant employers were caught for not paying the foreign employees their contractual fixed monthly salary or inflating the salaries with no intention of paying them the amount that has been declared. There were about 1,400 foreign employees affected in these cases.

(emphasis added by TWC2)

Do the second and third sentences truly respond to the question by Pritam Singh?

To know what the question was, watch this 3-minute video from Channel NewsAsia at this link. The key part of Pritam Singh’s speech was:

The Ministry of Manpower, arising from parliamentary questions filed by Workers Party MP Mr Faisal Manap and PSP NCMP Leong Mun Wai, recently confirmed that about 190 employers were found underpaying their foreign employees every year between 2015 and 2019. That total corresponds to close to one thousand employers over a five-year period. The total number of foreign workers who were underpaid was not disclosed, nor were any details of restitution made to these workers disclosed as well. Can the minister make these details known?

In the light of these words, look again at Tan See Leng’s response. Plenty of questions rain down.

  • What does “caught” mean?
  • How many of those “caught” were (a) prosecuted in court? (b) penalised through administrative measures?
  • Is the minister suggesting that over the five years 2015 – 2019, there were only 1,400 foreign workers who suffered salary shortfalls, and only 950 errant employers?

To be fair, we should bear in mind that this is in part an impromptu speech in a parliamentary debate. Perhaps the minister didn’t have the precise numbers that Pritam Singh was asking for. That said, it is interesting that the minister had the “950” and “1,400” numbers at hand, and not the more relevant numbers asked for.

Actually, the numbers are much larger

The Ministry of Manpower (MOM) and the Tripartite Alliance for Dispute Management (TADM), the latter which, since April 2017, has been dealing with employment disputes, have so far published two Employment Standards Reports. One edition was for 2017 -2018 (link), while the other was for 2019 – 2020 (link).

There are more numbers to be found there, though the way they are presented will make a reader tear his hair out. Different numbers in different sections refer to different time periods. Sometimes percentages are mentioned but not the base (or anchor) numbers. For incomplete years, sometimes the figures are annualised, sometimes not. And sometimes, not even clear which way they are.

This is what we can extract from the two published reports:

(If the figures are too small, you can click it to open the table in a separate window and you can then zoom it bigger)

Fortunately, however messy, there are enough numbers in those reports for us to interpolate among them to arrive at more consistently framed figures. It is important to remember that an interpolation exercise (especially when the original figures may suffer from rounding or definitional problems) does not yield exact figures. So, the figures in Table 2 should be treated as merely close estimates.

From Table 2, we can see that in 2018 and 2019 — the two complete years — foreign workers filed over 5,000 salary claims a year. This is far cry from Tan See Leng’s figure of 1,400 affected foreign workers over a five-year period.

Foreign worker salary claims made up over 60 percent of all salary claims filed at TADM during those two years.

As we all know, foreign workers do not make up over 60 percent of Singapore’s labour force, so the volume of their salary claims is disproportionate. There are statistics from MOM about the size of the local and foreign workforces, but it is also possible to derive from Table 1 itself  what those base numbers were.

Rows 27 to 32 of Table 2 show the derived estimates, which are pretty close to what MOM has published separately (see yellow box).

In short, foreign employees make up about a third of total employees, but filed over 60% of salary claims in 2018 and 2019.

As at 31 December 2018, MOM’s table (accessed 8 March 2021) showed 1,132,200 foreigners with various sorts of work passes, including Employment Passes (though the vast majority were Work Permit holders). As at 31 December 2019, the figure was 1,165,600.

From a different table, we see that the resident labour force (citizens and permanent residents) totalled 2,292,700 midyear 2018 and 2,328,500 midyear 2019.

Why are some figures mid-year and some figures end-year? Who knows?

An interesting factoid can also be found in a Channel NewsAsia story dated 11 October 2019 (link). It said:

Claims filed by foreign employees also involved larger salary arrears. The typical duration of unpaid wage claims by foreign employees was about two to 6.5 months, compared with the 0.5 to 2 months for locals.

Although the news story as a whole seemed to be reporting on the Employment Standards Report 2017-2018, your writer has not been able to find mention of this in the original report itself, so it is unclear where Channel NewsAsia got its information.

Success rate and restitution?

The Employment Standards Reports do not present enough numbers about success rate and restitution, nor any split between local and foreign employees on this score.

Instead, paragraph 9 on page 5 of the Employment Standards Report 2019-2020 says this:

Of the salary claims lodged between 1 January 2019 and 30 June 2020, 90% of employees successfully recovered their salaries fully at TADM or the Employment Claims Tribunals (ECT). About half of the remaining employees partially recovered their salaries. The total recovered sum amounted to about $23 million.

(emphasis added by TWC2)

And still we have questions!

Believe it or not, we don’t even know what “successfully recovered… fully” means. At TWC2, we come across many cases like these:

1. Reduction in claim amount under pressure
Example: a worker has been shortpaid $8,000 under various headings (basic salary, overtime, public holiday pay, annual leave equivalent, unfair deductions, etc). At the TADM mediation stage, he comes under pressure to jettison parts of his claim because the evidence he has is not very strong (e.g. employer refused to produce time cards that would prove his claim). He is pressured to reduce his claim amount to, say, $4,100. A settlement agreement is signed for $4,100 and the employer pays up. Has the worker “successfully recovered… fully” even though the shortfall in salary was $8,000?

2. Reduction in claim amount due to cap
Example: a worker has been shortpaid $27,000 and his evidence is strong. But under the Employment Claims Act, there is a cap of $20,000 on how much he can claim under this Act. He thus reduces his quantum of claim to $20,000 in order to avail himself of the process prescribed by the Act and eventually succeeds in his claim. The employer pays up $20,000 when so ordered by the Employment Claims Tribunal. Has the worker “successfully recovered… fully”?

3. Failure to honour Tribunal Order
Example: a worker has been shortpaid $13,000 and no settlement could be arrived at the TADM mediation stage. The claim goes up to the Employment Claims Tribunal where the magistrate finds in his favour. An Order of Tribunal is issued for the employer to pay the worker $13,000. No payment is made, and MOM is unable to enforce the order. MOM then tells the worker that as far as they are concerned, the case is “concluded”, and the worker is repatriated. Would such a case be counted among the “successfully recovered… fully” statistic simply because the tribunal has found in the worker’s favour?

What the above examples show is that reality is very granular. That being the case, it may be too much to expect a minister to reply in sufficient detail, especially if speaking off the cuff, but we certainly can expect the Tripartite Alliance to produce a better quality of report.

Just as company annual reports, for all their prose in the front pages, have to enclose detailed financial statements, complete with copious footnotes, so the Employment Standards Report should also contain detailed statistical tables and footnotes in its annexures. Nicely worded paragraphs and graphics — which one immediately associates with spin — do not substitute for the rigourous transparency and accountability that only statistical tables provide.