Why the Minimum Wage Harms the Economy?

In May 2011, Hong Kong adopted the minimum wage of $28 per hour. Employment in Hong Kong subsequently increased and the unemployment rate fell (see Table 1). Economic theory predicts the opposite should have happened if labor markets were competitive. Some commentators claim the outcome is evidence that either economics is wrong or the labor market is not competitive. So is economics indeed wrong?

 

Every first year economics course teaches that in a competitive labor market, the imposition of a minimum wage will cause employment to fall because when wages are raised the cost of doing business is higher and employers will hire fewer workers. The labor force participation rate and the unemployment rate will both rise in the labor market. This classic analysis is well known to any economics student. 

 

Monopsony Assumption Not General

 

Of course this analysis assumes that the labor market is competitive and employers do not have any monopoly power in the hiring of labor. A situation where employers do have market power is known as monopsony. When this happens, in theory it is possible for a minimum wage to increase employment as the job destruction powers of the minimum wage would not exist. But is this a likely situation, particularly for Hong Kong?

 

I don’t think so.

 

Minimum wage affects primarily unskilled labor. In Hong Kong, it is well known that workers change jobs instantly in response to very small changes in wages. This situation cannot be a sign of a labor market where employers are monopsonists. There is, however, a belief otherwise among some members of the intellectual public (see for example, 林行止, 最低工資不可廢 自我增值增入息, HKEJ, 2012.10.3).

 

This belief emerged after a number of highly regarded economists teaching at Berkeley, Princeton and the London School of Economics found that increases in the minimum wage did not decrease employment or increase unemployment among low paid workers, holding other things constant. Most economists have disputed these estimates as faulty and error ridden. A smaller, more sympathetic few conclude that the findings only hold for a certain geographical area, a particular industry, and a relatively short period of time. The best analysis to date on this line of academic inquiry is the recent work by Dube, Lester and Reich (2010) from Berkeley using US county level data.

 

Surge in Labour Demand

 

The inference drawn by these economists is that a minimum wage does not destroy jobs because there is monopsony power among some employers in the low paid segments of the labor market. Intuitively the argument goes something like this:

 

“Imagine a restaurant chain in a small city with high local commuting costs. Opening a second outlet might be profitable for the owner of the chain. However, he might have to pay higher wages to hire extra local staff in the vicinity of the new outlet. Since he is not able to pay different wages for staff between old and new outlets, this would drive up his labor costs in the old outlet. This effect harms the profitability of the old outlet more and might discourage the expansion of the business completely. However, if the government introduces a minimum wage, labor costs rise anyway and the prospective new outlet might become more attractive now. Hence the owner could open a new outlet and hire extra staff.”

 

But this line of argument relies on a very specific assumption about the size of the economy. It is premised on employers having monopsony power in a small city market in the presence of high local commuting costs. This feature renders the point made about the US experience to be of very limited relevance in real life Hong Kong. Hong Kong is one market in a big city economy, where retail chains have to compete against each other for labor. Commuting costs are quite homogeneous except for a few isolated areas. It is therefore not credible to suggest that retail chains currently possess monopsony power.

 

That situation may change in a few years when we will be able to tell more definitively whether retail chains in Hong Kong have monopsony power. If they do, then in the coming years the minimum wage will lead them to expand their outlets faster and hire more workers, while small and medium businesses in the same industry may contract due to higher wages. If not, then all enterprises – large chains and the small and medium businesses alike – will suffer employment losses or expand more slowly. Ironically, the minimum wage would make small and medium businesses less competitive especially when there is monopsony power among employers in the labor market.

 

I believe the employment outcome in 2011 reflected a very tight labor market situation in Hong Kong. Real consumption expenditure in the domestic market grew rapidly since 2010. Right at the time when the minimum wage was introduced the demand for labor extraordinarily high (see Table 1), which gave rise to an extraordinarily high demand for labor. The year-on-year rate of growth of consumption expenditure in fact peaked at 11.0% and 10.7% in 2011.Q2 and 2011.Q3.

 

Table 1: Labor Market Tightness and Domestic Consumption Growth Rate

								(%)
Year	Quarter	Employment Growth Rate		Labor Force Participation Rate		Unemployment Rate		Domestic Consumption Growth Rate
		Men	Women	Men	Women	Men	Women
2002		-2.8	1.2	72.4	51.9	8.3	5.9	0.8
2003		-1.7	0.1	72.0	51.7	9.3	6.2	-0.5
2004		2.1	3.3	71.6	51.9	7.9	5.6	7.7
2005		1.2	2.8	71.1	51.8	6.6	4.6	4.6
2006		1.0	3.1	70.9	52.6	5.7	3.8	6.2
2007		1.3	3.3	70.4	53.1	4.7	3.4	9.7
2008		0.1	1.9	69.7	53.1	3.9	2.9	2.6
2009		-2.1	-0.1	69.4	53.2	6.0	4.3	1.4
2010		0.4	-0.1	68.5	51.9	5.1	3.6	8.8
2011		1.7	4.3	68.4	53.0	4.0	2.8	9.7
2008	Q1	1.1	3.8	69.7	53.1	3.6	2.6	8.1
	Q2	0.6	2.3	69.5	53.0	3.8	2.8	4.8
	Q3	0.3	1.7	69.8	53.2	4.0	3.0	1.0
	Q4	-1.4	1.6	69.5	53.2	4.2	3.2	-2.7
2009	Q1	-2.1	0.7	69.5	53.7	5.6	3.7	-3.2
	Q2	-2.0	0.4	69.7	53.5	6.3	4.4	0.0
	Q3	-1.9	-0.3	69.7	53.2	6.6	4.9	2.0
	Q4	-1.8	-0.4	68.5	52.6	5.7	4.1	6.8
2010	Q1	-0.2	-0.4	68.5	52.1	5.2	3.6	9.6
	Q2	-0.2	-0.2	68.3	52.1	5.6	3.6	7.3
	Q3	0.9	0.3	68.7	51.8	5.1	3.8	8.9
	Q4	1.3	0.6	68.5	52.0	4.5	3.3	9.4
2011	Q1	0.3	1.9	67.9	52.5	4.1	2.6	8.7
	Q2	1.7	3.3	68.5	53.0	4.3	3.0	11.0
	Q3	1.3	4.4	68.6	53.2	4.0	3.0	10.7
	Q4	1.2	4.9	68.2	53.4	3.7	2.7	8.4
2012	Q1	2.7	3.9	68.9	53.6	3.8	2.5	7.0
	Q2	2.1	4.0	68.8	54.1	3.8	2.8	4.4
	Q3	1.1	2.2	68.8	53.6	3.8	3.1	3.0

                                                                               

The increase in labor demand resulted in a net increase in employment. Total employment of men and women in Hong Kong has been increasing since the second half of 2010. The higher demand also increased the labor force participation rate and decreased the unemployment rate. The labor force participation rates for men and women have risen from 68.5% to 68.8% and from 52.1% to 53.6%, respectively, over the period 2010.Q1-2012.Q3. The unemployment rates for men and women have fallen from 5.2% to 3.8% and from 3.6% to 3.1%, respectively, over the same period. Most of the negative job destruction effects of a minimum wage thus were concealed, for the time being, by introducing the wage in a tight labor market.

 

Harms the Overall Economy

 

Does this mean that minimum wages can do no harm if the labor market is tight? To answer this question let us consider an example. We assume that there are three types of low skilled jobs in the Hong Kong labor market, X, Y and Z, paying $30, $28 and $26 before May 2011. Let us assume there are two types of workers who are hired by employers operating in this segment of the labor market. Worker type A+ is more enthusiastic. Worker type A- is less enthusiastic. The difference in their productivity is largely a result of the difference in their enthusiasm and not ability. Let us assume there are an equal number of A+ and A- workers in the labor market. What is a likely possible allocation of A+ and A- workers to these three types of jobs? It would depend on the ability of employers to recruit and retain enthusiastic workers.

 

Table 2 gives a possible allocation of workers among jobs before and after the minimum wage was introduced. Before the minimum wage was introduced we hypothesize there are two A+ workers and 6 A- workers in Job X; four A+ and four A- workers in Job Y; and six A+ and two A- workers in Job Z. Since A+ workers are the more enthusiastic workers, therefore the productivity of workers in Job X is on average lower than Job Y, and Job Y is on average lower than Job Z. Naturally Job Z will have to pay more than Job Y, and Job Y will have to pay more than Job X. Employers in Job Z have been more successful in recruiting and retaining enthusiastic A+ workers and have rewarded their workers on average better than workers in Job Y. Employers in Job X have been least successful in recruiting and retaining enthusiastic A+ workers and have ended up paying their workers least well on average.

 

Table 2: Wage Changes and Allocation of A+ and A- Workers by Jobs

		Job X	Job Y	Job Z
Before minimum wage is introduced	Wage rates	$26	$28	$30
	Number of type A+ workers	2	4	6
	Number of type A- workers	6	4	2
Step 1 after minimum wage of $28 is introduced	Wage rates	$28	$28	$30
	Number of type A+ workers	4	2	6
	Number of type A- workers	4	6	2
Step 2 after minimum wage of $28 is introduced	Wage rates	$28	$30	$30
	Number of type A+ workers	4	4	4
	Number of type A- workers	4	4	4

 

 

 

After a minimum wage of $28 is introduced, workers in Job X are paid the same wage rate as workers in Job Y. The labor market will respond in two ways. In Step 1, enthusiastic A+ workers will become indifferent to working in Job X or Job Y since the pay is the same. It is conceivable some A+ workers in Job Y will now have an incentive to shift to Job X, which previously did not reward enthusiastic workers well but at the same time was more tolerant of less enthusiastic workers. Suppose two A+ workers finds shifting from Job Y to Job X preferable. Job Y now suffers a loss of two enthusiastic A+ workers. Its market performance will suffer and Job Y loses competitiveness in the market relative to Job X. The minimum wage therefore affects different jobs differently; it is not neutral in its effects on industry.

 

Job X becomes more productive with an addition of two A+ workers to replace two A- workers. This is consistent with some research observations about possible productivity enhancing effects of an increase in the minimum wage. But it is important to note that the productivity gains in Job X occur due to a reshuffling of enthusiastic and less enthusiastic workers across jobs and not because there is any real productivity gain in the economy.

 

Since Job X could previously survive with two less enthusiastic A- workers. therefore they may fail to fully utilize the enthusiasm of the two new A+ workers. Enthusiastic A+ workers are now placed in a job where tolerance for being less enthusiastic is greater. If that is the case then the gain of two A+ workers, although beneficial for Job X, is in effect a loss to the economy as a whole.

 

Step 1 of the adjustment in the labor reveals two detrimental effects on the economy. First, Job Y becomes less productive. Second, Job X does not gain as much productivity as is possible. Enthusiastic workers end up in the industry that values enthusiasm less, and vice versa.

 

We now consider Step 2 in the adjustment to a minimum wage. In a bid to restore competitiveness in Job Y, the employers raise their wages to attract enthusiastic A+ workers. To do so they have to raise wages to $30 in the market. Let us suppose they are successful in recruiting two A+ workers from Job Z and replace two of their A- workers. If this happens then Job Z ends up with four A+ and four A- workers. The profile of their workers is worsened and they suffer a loss of productivity. Job Z now loses competitiveness relative to both Job Y and Job X.

 

Labour Races to the Bottom

 

We can of course continue to have Step 3 and more, whereby wage increases percolate up to other jobs and enthusiastic workers move down into jobs that do not require such workers. But the story is clear. Minimum wages change the relative competitiveness of workers across industries and move enthusiastic workers down the job ladder. This misallocation of workers appears because wage rates near the minimum wage level provide a distorted signal for the efficient allocation of workers. This harms the economy.

 

In the past year we have witnessed this happening after minimum wages have been introduced. Housing estate guards and watchmen are becoming younger. Restaurant servers in prime central areas are experiencing labor market tightness.

 

Compare now the allocation of A+ and A- workers among the three jobs when Step 2 has been reached. There are four A+ and four A- workers in every job. Such an allocation is very different from the original allocation of workers before the minimum wage was introduced where enthusiastic workers were more concentrated in the high paying Job Z and less enthusiastic A- workers were mostly found in low paying Job X. The random spreading of different types of workers into jobs that are different damages the market system. It homogenizes differences that are present and should be present. The market provides a mechanism to sort people into different jobs. The minimum wage destroys this important function for low paying jobs.

 

Increasing the minimum wage coverage to more workers would further enlarge the damage to the economy. It simply creates a race to the bottom for a larger number of enthusiastic workers.

 

References:

 

Arindrajit Dube, William Lester and Michael Reich, “Minimum Wage Effects Across State Borders: Estimates Using Contiguous Counties”, Review of Economics and Statistics, Vol.92, Issue 4, 2010.