Calculating productivity: The devil is in the statistics
MODERATING the influx of foreign labour is one of the necessary steps in helping to achieve the aim of raising labour productivity by 2 per cent to 3 per cent annually on average during this decade ("Firms welcome clarity on foreign labour policy"; last Thursday).
The method of dividing gross domestic product by total employment has been used by many countries to obtain a quick assessment of labour productivity.
The method is crude, and for a small and very dynamic, fast-changing economy like ours, it poses problems for serious productivity studies and policymaking.
For example, a big jump in pharmaceutical production could affect the productivity of the whole manufacturing industry significantly. Unless details are given, one is unable to know how other parts of the manufacturing industry are doing.
Another example: When two partners of a small eatery plan to retire in one or two years, and hire one foreigner to help them, the number of workers in the eatery would immediately increase by 50 per cent.
If another part-time local worker is employed later and counted as one full worker, the employment figure would double and the value-add per worker would be halved if business does not improve very much.
Unless we keep updated records of retirements and semi-retirements, and make adjustments for part-time workers, the derived productivity statistics could be misleading.
Our labour shortage problem is not going to ease, as witnessed by the rapid rise in job vacancies from about 19,000 in 2005 to 52,000 last year.
To facilitate and enhance the planning, implementing and refining of manpower policies, it is important that we improve our productivity statistics.
Detailed productivity information will also give us a more accurate picture of how we fare in our productivity drive.
Ng Ya Ken