Much of the expense to develop English language competence is born by the government, as a result employers necessarily utilize more English than they actually require. There are several points to consider in this regard.

• Social efficiency
  hypothesis testing (index)
  
  Economic theory suggests that the optimal level of English language training in society as a whole is reached when the marginal social benefit of using English is equal to the marginal social cost of learning and maintaining it. Although it is difficult to measure these social costs and benefits directly, a positive rate of attrition for a population as a whole would suggest an inefficient use of resources.
  
  Moreover, since there are many schools in the private sector that provide additional sources of training, an average positive rate of attrition for the entire population would suggest that not only does government provide too much English language training, but that which it does provide is inappropriate.
  
  
• Excess demand
  hypothesis testing (index)
  
  If government is willing to provide a valuable resource at no cost to employers, then employers are likely to use far more of the resource than that which they actually require.
  
  Firms tend to use a resource until the marginal benefit obtained from its use equals the marginal cost of its further employment. If the marginal cost of the resource is zero, firms pay nothing for the resource, and employ it to the point where it is no longer of benefit. Government supplies industry with low level English competence at no cost to industry, but at substantial cost to the general public.
  
  When the government asks firms to report what they employ, they necessarily report an amount far higher than that which they actually require. This is because what they employ is based on a marginal cost of employment that does not reflect the true value of the resource. To put it another way, firms would employ far less of the English language, if they were made to pay the cost of producing it. Moreover, they would likely find other means to satisfy the same ends, at a cost far more in line with the true market value of the English language.
  
  
• Accurate assessment
  hypothesis testing (index)
  
  Under the assumption that the government can provide English language training to those who require it at least as efficiently as the private sector, it is still important to determine exactly how much training the private sector actually requires. In order to achieve this one must be able to distinguish employment with and without government supplied language training.
  
  For the reasons already given a survey of current business needs based upon current English language usage would not provide an appropriate assessment of Hong Kong's English language requirements. A more appropriate assessment would seek to determine the amount of language employed, if business and government were made to pay the full cost of English language training.
  
  
• Nature of the market
  hypothesis testing (index) | perceived use vs actual need)
  
  Although government and industry are essentially provided with a "free" language asset, the quality of that asset varies from individual to individual. Moreover, not all jobs require the same level of language competency. Language competency is a measurable asset, and government and business are thus able to discern different levels of competency among potential job candidates and their employees. As such, both employers and employees will exploit these differences to their own advantage. Employers with important language needs use wage incentives and employee benefits to attract and maintain those employees with the highest levels of competence. Similarly, job candidates and employees use their language competency as a bargaining tool to obtain higher wages and better working conditions -- in short, the best employment possible. Thus, despite the rudimentary level of competency that government provides itself and business, a market has developed in order to exploit varying levels of competency and need.
  
  Because a market for language competence exists it is reasonable to believe that the market worth of government language training can be estimated, and its rate of return as an investment in human capital measured.
  
  
• Return on investment
  hypothesis testing (index)
  
  By asking government and private sector employers how much they would pay for incremental differences in language competence among their employees, one can estimate the productive worth of various levels of language competence to Hong Kong employers. Dividing these incremental differences into the amount spent by government on English language education for each of its citizens, one can determine the government's relative rate of return on investment for each level of language competence that it produces.
  
  From these estimates one can better determine how government money should be spent on universal English language education, if indeed, it should even be required.
  
  Though one would expect higher rates of return from higher levels of competence, this may not be the case. This is because those who pay for the cost of asset development (government) and those who purchase and utilize the developed assets (government and business) are not the same. Moreover, a too wide disparity in relative rates of return between low and high levels of English language competence would suggest that less emphasis should be placed on universal language training and more on high quality language competence.