Response to the study report on 'Minimum Wage Systems'
compiled by the Legislative Council Secretariat


At its meeting on 27 May 1999, the Legislative Council Panel on Manpower discussed a report on "Minimum Wage Systems" compiled by the Research and Library Services Division (RLSD) of the Legislative Council Secretariat. During the discussion, the Panel requested the Administration to render a response to the RLSD report. This paper sets out the Administration's response to the report.

General comment on the RLSD report

2. The RLSD report contains a substantial amount of information on the experience and practices of a number of overseas economies in implementing their minimum wage systems, covering such aspects as historical development and evolution, enforcement mechanism, social and economic impact assessment, etc. The information is presented in a well organised and factual and objective manner. On the effect of changes in minimum wage on employment, poverty situation and income distribution in the respective economies, the pros and cons are laid out in a balanced manner. Understandably, for the sake of neutrality, the RLSD report has not gone so far as to conclude whether or not Hong Kong should set up a statutory minimum wage system.

3. There are however some key aspects which the RLSD report does not seem to have explicitly addressed. In drawing observations from the various studies on overseas economies, there is little mention of the more recent trend towards liberalisation of the minimum wage system particularly in the advanced economies, as part of their move to reduce rigidities in their labour market. Also, the empirical studies have all focused on the implications arising from a change in the statutory minimum wage level, but not the implications of implementing statutory minimum wage as a brand-new system, which should be of more direct relevance to Hong Kong. Furthermore, the RLSD report has not made sufficient reference to a comprehensive research study on the subject done recently by the Organisation for Economic Cooperation and Development (OECD), which covers many more economies and provides some overall observations(1).

4. In this response, we seek to address these key aspects. In addition, we examine broadly the evolution of the conceptual argumentation pertaining to statutory minimum wage up till recently. Then we turn to the likely implications of imposition of statutory minimum wage, at say about $4,000 - $7,000 per month, for the Hong Kong economy. Concluding this response, we re-iterate the Administration's stance on the issue, i.e. statutory minimum wage should not be introduced in Hong Kong.

Evolution of conceptual argumentation pertaining to statutory minimum wage

5. The long-standing mainstream argumentation against imposition of statutory minimum wage can be summarised as follows. First, for the covered sectors, the loss in employment thus resulted could outweigh the gains by those minimum wage workers still in employment. Secondly, unless the low-pay workers belong to the poor households, the statutory minimum wage might end up benefiting low-wage members of better-off households. Thirdly, the surplus labour generated from the covered sectors could depress wages in the uncovered sectors, thereby offsetting any poverty alleviation gains in the covered sectors. If the whole economy were covered, higher overall unemployment would result.

6. In recent years, some empirical studies have tried to resurrect the case for statutory minimum wage, by alluding to economic efficiency instead of poverty alleviation grounds. A modern theory of "dynamic monopsony" was advanced. As distinct from the conventional theory, it does not rely on the extreme case of "company town" where the sole employer company is able to exercise monopsonistic powers in setting wages for its workers at below the market-determined level. Rather, the idea is that, in a labour market where many small employers exist, workers will find it less easy to obtain complete information about wage level and dispersion than in a labour market dominated by a few large employers. It is argued that, in the light of such information deficiency, employers may also be able to set wages at below the market-clearing level. It is therefore claimed that, under such circumstances, imposition of a statutory minimum wage at above the prevailing level would actually improve economic efficiency.

7. But still more recently, quite a number of economists have made strong rebuttals. They consider that, as employers and employees adapt to uncertainties in the labour market, some divergence every now and then between wage and the marginal product of labour is inevitable. But the existence of such divergence in practice, whereby the dynamic monopsony theory comes out as a means of rationalisation, does not really provide a sufficient case for accepting the clear distortion brought about by a statutory minimum wage. Moreover, competition for labour from the numerous employers will sustain a tendency to bid up wages to the market-clearing level. On the other hand, statutory minimum wage will have the undesirable effect of reducing incentive amongst the workers to acquire skills. This could damage the dynamic efficiency of the economy. (For more details, see Annex I.)

Economic and social implications of statutory minimum wage for the OECD economies

(a) Employment opportunities

8. According to the OECD study, in order for a statutory minimum wage system to be effective, the statutory minimum wage should be set at some margin above the market-clearing level. However, if so, this would lead to a surge in unemployment, especially amongst those workers whose productivity is not sufficient to match the statutory minimum wage. The extent of job loss would tend to vary amongst different economies, depending on their respective economic and labour market characteristics(2). Generally speaking, the loss in employment would be more apparent over the longer term, as employers could have sufficient time to adjust their employment practices so as to minimise the entailing cost impact(3).

9. Several other studies also come up with similar findings. In particular, a study by the National Bureau of Economic Research in the United States has found that a 10% rise in the statutory minimum wage in real terms would lower the probability of future employment by 10-13% in France, and by 4 - 16% in the United States(4).

(b) Protection of low-wage workers

10. Ironically, the increase in unemployment resulting from statutory minimum wage is found to concentrate in the less-advantaged workers, including in particular teenagers, disabled persons, unskilled workers etc, who normally receive low pay and thus are intended to be the primary beneficiaries of the statutory minimum wage system. Yet owing to the relatively lower skill and productivity for many of these workers, they are often out-competed by the others in the job market. This observation is made in the RLSD report(5). The OECD study likewise reveals that the reduction in employment tends to be more visible amongst teenage workers aged 15-19. It is estimated that a 10% increase in the statutory minimum wage in money terms would result in a reduction in teenage employment by 2-4%. A separate study by the National Bureau of Economic Research(6) indicates that the impact would be most significant amongst the "marginal workers". These include basically men aged 25 - 35 whose wage just prior to the increase in statutory minimum wage is above the previous minimum but below the new higher level. According to the study, an upward adjustment of 1% in the statutory minimum wage is expected to reduce the probability of keeping the jobs of such marginal workers by 4.6%.

11. Even for those low-wage workers who manage to retain their jobs after implementation of statutory minimum wage, their overall remuneration against workload might not increase eventually, as employers would seek to circumvent the additional cost burden either by trimming their non-wage benefits and/or by increasing their workload. In the end, the statutory minimum wage system might not provide much relief to the plight of low-wage workers.

(c) Training for better skills and productivity

12. In face of increased labour cost upon implementation of statutory minimum wage, employers would become less willing to provide training for their workers(7). At the same time, the low-wage workers themselves would have less incentive to pursue further training and retraining so as to sharpen their skills in their career paths. More specifically for teenage workers, the reduced job opportunities consequential to implementation of statutory minimum wage would retard the process of on-the-job training for better skills and productivity, to the disadvantage of the employers, the employees themselves, and the economy at large.

13. The remuneration packages for employees are also likely to be altered upon the implementation of statutory minimum wage, by trimming the variable, performance-oriented rewards such as bonuses and commissions so as to make up for the statutory minimum wage as a rigid, fixed pay component. This in turn would dampen work incentive and hence productivity performance of the workers concerned. The impact is expected to be greater on the service sectors, given the greater importance of performance-related rewards in the overall remuneration packages of employees in those sectors.

(d) Economic adjustment process

14. At a time of economic downswing, it is essential for labour wages to be able to adjust downward in a flexible and expeditious manner. This will help reduce production costs and hence output prices thereby enhancing competitiveness, and so is conducive to revival in overall demand. However, judging from the experience in the OECD economies, statutory minimum wage is generally rigid downward, though flexible upward. This downward stickiness in wage movement hampers the adjustment process in both the labour and the product markets during an economic downswing, thereby prolonging the unemployment situation and retarding recovery of the economy. Worse still, with an entrenched downward rigidity in the wage adjustment mechanism, what originates as shorter-term cyclical unemployment might persist into long-term structural unemployment.

(e) Restructuring of the economy

15. When the economy is undergoing structural change, it is crucial for wages to be determined to the maximum extent possible by market forces in an uninhibited way, so as to ensure an efficient allocation of manpower resources amongst the various economic sectors as the pattern of labour demand progressively changes. In response to the market signals provided by the changes in relative wages, workers will shift from sectors of declining importance to those of growing significance. A statutory minimum wage would blur the pattern of relative wages and its necessary change in line with economic restructuring, thereby retarding the flow of manpower resources and hence the very process of restructuring for better gains to the entire economy.

(f) Alleviation of household poverty

16. Contrary to common perception, there is no close relationship between minimum wage and poverty reduction in an economy, for the following reasons. First, for those low-wage workers who are laid off upon the implementation of statutory minimum wage, they are deprived of job opportunities and hence employment earnings. Secondly, for those low-wage workers who manage to retain their jobs, the increase in wage up to the statutory minimum will have to be weighed against the possible cuts in non-wage benefits and allowances by their employers. Thirdly, statutory minimum wage cannot improve the livelihood of those poor households comprising mainly the unemployed or the economically inactive persons such as retirees, housewives and children, as households with no working members cannot really benefit from such a provision. A congressional study in the United States also comes up with this observation(8).

17. Statutory minimum wage may help to reduce household poverty only if there is a high concentration of low-wage workers in poor households. Yet the OECD study reveals that, in the United Kingdom and the Netherlands, only 5% and 13% respectively of the adult members in poor households are low-wage workers, while the corresponding figure for the United States, though considerably higher at 33%, is not so large as to be able to equate low wage with poverty(9) (Annex II). The OECD study thus concludes that statutory minimum wage is far from an effective solution to the poverty problem in the economies concerned.

18. Summing up, the experience of the OECD economies shows that implementation of statutory minimum wage generally cannot attain the desired objective of helping to alleviate poverty. As most of the poor households do not have working members, minimum wage cannot provide the assistance intended. Even for those households with low-wage members, the threat of job loss puts them in a more worrisome position. Moreover, at the macro level, a statutory minimum wage system gives rise to rigidity in the labour market structure, resulting in higher labour cost, lower productivity, weaker international competitiveness, and reduced job creation. Learning from their past mistakes, many of the OECD economies have turned to liberalisation of the labour market in order to re-invigorate their economic dynamism.

General trend towards relaxation of the statutory minimum wage system amongst OECD economies

19. There has manifestly been a general trend amongst the OECD economies to introduce greater flexibility into their statutory minimum wage systems(10), particularly in regard to such specific categories as youth workers, disabled workers, and apprentices and trainees. For instance, in the United States, a separate minimum wage rate at a lower level has been fixed for workers aged below 20 since 1996. Also, upon fulfilment of certain conditions, employers can pay 85% of the statutory minimum wage for trainees, as well as lower rates for disabled workers. In New Zealand, the statutory minimum wage rate for youth workers has been pitched at 60% of the corresponding statutory minimum wage rate for adults since 1994. Likewise, in the Netherlands, Belgium and Luxembourg, the statutory minimum wage rate for workers aged below 23 is generally lower than those for workers at the higher age groups.

20. In addition, some OECD economies have sought to offset the higher wage cost to employers brought about by the statutory minimum wage, by reducing the statutory non-wage items for low-wage workers(11). In France, employer's social security charges have been reduced on two occasions since 1993, with the effect that employer's social security contributions for the minimum-wage workers are lowered by 40%. In Belgium, employer's social security charges were similarly reduced, by up to 50% for the low-salary workers in 1994. In the Netherlands, employer's payroll tax for the lower-pay workers was reduced, by some 25% in 1996.

21. Over the years, several OECD economies(12) have also strived to do away with the practice of indexing the statutory minimum wage automatically to inflation, by replacing it with an adjustment mechanism based on a wider range of market-related factors including competitiveness, business viability and employment situation. Partly due to this change in practice, statutory minimum wage in the OECD economies has generally risen at a slower pace than overall wage growth. As an indication, the ratio of statutory minimum wage to median earnings in many of the OECD economies, including the United States, Canada, New Zealand, the Netherlands, Mexico, Portugal, Greece, Turkey and Spain, has followed a declining trend over the past 25 years, falling from a range of 50-80% in the early 1970s to that of 20-50% in 1996(13).

Implications of statutory minimum wage for the Hong Kong economy

22. By reference to the experience in some selected overseas countries, our inference from the RLSD report is that, were the minimum wage system to be implemented in Hong Kong, the statutory minimum wage might be set at around HK$4,000 - 7,000 per month(14). Some analysis on the likely implications thus arising is provided below.

(a) Elevating unemployment

23. If the statutory minimum wage were set at HK$4,000, some 100 000 workers or 3% of the local workforce, excluding foreign domestic helpers, would be captured(15) (Annex III). But if it were set at HK$7,000, the corresponding figures would be much greater, at 500 000 and 16%. Amongst them, marginal workers with lower skills and productivity can be expected to face greater risks of losing their jobs upon implementation of statutory minimum wage. Female workers and workers in the unskilled category are likely to be affected to a greater extent (Annex IV(a)-(b)).

24. The effect on unemployment in Hong Kong brought about by implementation of statutory minimum wage could be significant for the following reasons. First, given a flexible labour market, the wage elasticity of both labour demand and labour supply in Hong Kong are unlikely to be low(16). This means that the reduction in employment caused by a specific administrative lift in the wage rate would probably not be small. Secondly, part-time and casual workers would be more seriously affected, if the statutory minimum wage is also rigid in regard to working hours. Thirdly, as sounded out so far, the objection from the local business community to the implementation of statutory minimum wage is strong. A likely reaction in the event that this measure is imposed would be that more of their business operations, along with the jobs provided, would move out of the territory.

(b) Aggravating the plight of workers

25. Statutory minimum wage, if implemented, would exacerbate the already severe unemployment rate faced by teenagers. In the first quarter of 1999, the unemployment rate for persons aged 15-19 was estimated at around 23%, way above that of 9% for persons aged 20-24 and of 6.2% for the entire labour force (Annex V). Nearly three-quarters of the workers in the teenage group were at the semi-skilled and unskilled levels, as compared to 60% for the entire labour force (Annex VI).

(c) Rendering insignificant poverty alleviation effect

26. Following the methodology adopted in the OECD study, it is estimated that only 12% of the adult members living in the poor households are low-wage workers (Annex II). Within these poor households, about three-fifths of the adult members are economically inactive or outside the working population, comprising mainly retirees, housewives and students, and also nearly one-tenth of the adult members are unemployed (Annex VII). Obviously, those households without working members cannot benefit from the statutory minimum wage. Coupled with possible job loss for those low-wage workers in the poor households upon implementation of statutory minimum wage, it is rather unlikely that the poverty faced by such households would be much relieved as a result. In any event, the social safety net accorded by the Comprehensive Social Security Assistance Scheme (CSSA) has been performing its proper function of rendering assistance to those households in genuine financial hardship.

(d) Prolonging economic downswing

27. As of now, the Hong Kong economy is still undergoing extensive adjustment in the wake of the regional financial turmoil. In the process of this adjustment, the economy shows a high degree of flexibility in respect of prices and rentals, as manifested by markedly lower property prices and rentals and prevalence of price discounts at retail outlets, and this is helpful for preserving international competitiveness. By comparison, nominal wages, though also following a moderating trend, seem to have come down less distinctly. Any move to introduce greater rigidity into the wage mechanism, as would be so by implementing statutory minimum wage, is bound to impede the adjustment process. This in turn would prolong the cyclical rise in unemployment and retard recovery of the economy at large.

28. It should be noted that, even with a pick-up in economic activity, employers could still be cautious about hiring additional workers, at least not so soon after they have just effected downsizing. Experience in the OECD economies indicates that a "jobless recovery" in the early stage of upswing is quite plausible(17). Implementation of statutory minimum wage at that stage would certainly not help resurrect employment.

(e) Adding to cost burden for businesses

29. On cost burden, the more labour-intensive businesses with a larger proportion of lower-skill workers at lower pay would be affected more than the others by the implementation of statutory minimum wage. Payroll-related costs account for a large proportion of the total operating costs in such sectors as restaurants, retail trade, security services, and repair, laundry and domestic services (Annex VIII). The impact would tend to be more pronounced on small to medium-sized enterprises, many of which probably have less lattitude than the large enterprises in adapting to the cost burden thus imposed.


30. In conclusion, we would like to make the following points:

  1. The RLSD report has considerable breadth, with a lot of factual information on the experience and practices in a number of economies in implementing statutory minimum wage systems. Yet after going through the report, as well as canvassing other references, we maintain that there is no case to change our policy stance on the issue.

  2. The contents of the RLSD report do not provide any strong grounds on the whole for setting up a statutory minimum wage system in Hong Kong. At best, the economic effects are shown to be mixed. The adverse consequences thus entailed cannot be lightly taken.

  3. Of particular concern is the likelihood that implementation of statutory minimum wage would impinge most on the jobs of the marginal workers and the low-wage workers, contrary to the desire that statutory minimum wage would be able to protect or benefit them.

  4. While some empirical studies suggest that statutory minimum wage may bring about little negative impact, or even some positive impact, on employment, such studies are often found to be fraught with analytical problems. Most of the other studies, including notably that conducted by OECD, reaffirm the view that statutory minimum wage could have a negative impact on employment and may not help much in alleviating poverty.

  5. The institutional set-up and operational mechanism involved for establishing and administering a statutory minimum wage system are likely to be elaborate, and moreover are prone to conflict. The RLSD report shows that for implementing a statutory minimum wage system, there will have to be layers of boards and committees with wide representation from the employers, trade unions, government officials and academia. Pursuance of any change in the statutory minimum wage will be very time- and resource-consuming. Not least are the host of statistical indicators, including wage level and differentials, operating costs of businesses, competitiveness of the economy, prices, employment and so on, that the bureaucratic organ concerned will have to go through in order to seek agreement amongst the divergent vested interests.

  6. Even Singapore, where the government has much greater direct involvement in the workings of the economy, has not implemented a statutory minimum wage system. The RLSD report mentions that the Singapore government is staunchly against statutory minimum wage on the ground that it would introduce inflexibility into the wage system. It postulates that wages should be determined by market forces of supply and demand. Employers and workers should be allowed to negotiate and mutually agree on the wages to be paid before they enter into employment contract. The role of the National Wage Council is rather to enhance Singapore's competitiveness in the international market by adjusting the wages, either upward or downward as necessary, in an orderly manner so as to foster employment. Moreover, more recently it places strong emphasis on enlarging the variable component of the pay package for the sake of reducing wage rigidity.

  7. The market-determined wage setting mechanism in Hong Kong is well-tried and time-tested. In no small part, the smooth operation of market forces has enabled the Hong Kong economy to attain robust growth over the past decades, resulting in a substantial improvement in standard of living for the community. It is of utmost importance for the economy to maintain resilience and flexibility in both the product and the factor markets, including the labour market, for economic growth to be sustained over the longer term.

  8. The following statement from an article by Deepak Lal(18) is illuminating, as an ending remark:
"The minimum wage is an inefficient, well-intentioned but inexpert interference with the mechanisms of supply and demand."

So Hong Kong should not fall into this pit.

Economic Analysis Division
Financial Services Bureau
Government Secretariat
17 July 1999

(1) OECD (1998), "Making the most of the minimum: statutory minimum wages, employment and poverty", in OECD Employment Outlook, June 1998.

(2) The factors that would come into play include the specific differential between the statutory minimum wage and market wage, elasticity of demand for and supply of labour, degree of substitution between workers of different skills, etc.

(3) Market conditions permitting, employers would shift the cost burden to consumers through a rise in product prices. Alternatively, they could resort to other means such as production rescheduling, corporate downsizing, closure of marginal activities, and in the extreme case, cessation of the entire business. In most of these cases, a reduction in employment opportunities would result. The experience of the United States is a notable case in point. In response to the hike in statutory minimum wage in the late 1970s and again in the early 1990s, many restaurants in the country reportedly had eliminated table service, promoted self-service, and substituted machinery for labour in food preparation, while some of them cut their business hours short. All these moves led to reduction in jobs.

(4) Abowd, Kramarz, and Margolis (1999), "Minimum wages and employment in France and the United States", National Bureau of Economic Research, Working Paper No. 6996, March 1999.

(5) See the last sentence in paragraph 9.19 of the RLSD report, which reads "It is generally agreed that a high real minimum wage could have an adverse impact on the employment of low-wage workers."

(6) Abowd, Kramarz, Lemieux and Margolis (1998), "Minimum wages and youth employment in France", National Bureau of Economic Research, 1998.

(7) Neumark and Wascher (1998), "Minimum wages and training revisited", National Bureau of Economic Research Working Paper, No. W6651, July 1998.

(8) McCallion (1996), "The Federal Minimum Wage: Employment and Distributional Effects", Congressional Research Service Report for Congress, May 1996. Specifically, the reasons cited in the report are: (i) not all minimum wage earners live in poor households; (ii) most of the working poor already earn more than the minimum wage; and (iii) some poor households do not have wage earners.

(9) In the OECD study, "poverty" is defined in the relative sense, referring to those households having per capita income levels less than half of the median per capita household income. As to low-wage workers, they refer to those earning less than two-thirds of the median employment earnings.

(10) A distinct exception is the reinstatement of statutory minimum wage in the United Kingdom by the Labour Government in April 1999, following its abolition by the Conservative Government in 1995. Yet arguably, this merely reflects the change in political orientation in the United Kingdom itself.

(11) OECD (1997), Implementing the OECD Jobs Strategy : Member Countries' Experience, pages 72-75.

(12) Notable examples are Greece, Italy and the Netherlands.

(13) OECD (1998), "Making the most of the minimum: statutory minimum wages, employment and poverty" in OECD Employment Outlook, June 1998, pages 40-41.

(14) In Part 13 of the RLSD report, the ratio of minimum wage to average wage is found to be about 60% in France, 45% in the United States, the United Kingdom, Taiwan and Japan, and below 40% in South Korea. Applying a benchmark ratio of 40-60% thus inferred to the average wage level in Hong Kong, the hourly minimum wage for consideration might range from HK$23.7 to HK$35.6, and the monthly minimum wage from HK$4,554 to HK$6,830 per month. For the purpose of this analysis, we broaden the latter range slightly to HK$4,000 - 7,000 per month.

(15) As at end-March 1999, there were around 180 000 foreign domestic helpers in Hong Kong. They are subject to a Minimum Allowable Wage.

(16) As an indication, the wage elasticity for the United States, Germany, the United Kingdom, France and Japan are estimated at 0.8 - 1.0, according to an OECD study. Our estimate suggests that the wage elasticity for Hong Kong falls within this range.

(17) As an illustration, the United States economy bottomed out from recession to resume positive GDP growth in the second quarter of 1991, and has undergone continuous growth since then. Yet its unemployment rate still rose further in the ensuing quarters, from 6.8% in the second quarter of 1991 to a peak of 7.6% in the third quarter of 1992. For details, see OECD (1994), "The recent recession and current recovery in historical perspective", in Part I of The OECD Jobs Study : Evidence and Explanations.

(18) Deepak Lal (1995), "The Minimum Wage: No Way to Help the Poor", IEA Occasional Paper 1995.