But we know that governments do not create jobs - the Prime Minister and Treasurer have become quite careful on this point, while still boasting that there will be 500,000 extra jobs during the next two years - which is, by the way, a slowing in the rate of employment growth.

In the May budget, the government has managed to find $715 million over four years to fund a skills package, although there are some previously committed training initiatives that were ditched as part of the deal. The Critical Skills Investment Fund, announced only last year, is no longer and Productivity Places, the much maligned and ineffective training program, has been ditched. In their place are a series of new programs, including:

► The $558m National Workforce Development Fund, which will deliver 130,000 new training places in four years.

► $101m for a national mentoring program to help 40,000 apprentices finish training.

► Accelerating apprenticeships by investing $100m in more flexible training models.

► 30,000 more places in the Language, Literacy and Numeracy Program to provide the basic skills essential for a job.

► $233m in new support programs and 35,000 targeted wage subsidies to encourage employers to hire those who have not worked for more than two years.

There will be a new agency, the Orwellian-titled National Workforce and Productivity Agency, which will undertake workforce planning.

The agency will have a board made up of representatives of employer associations, industry skills councils and trade unions. This agency effectively takes over the functions of Skills Australia, which hilariously informed us recently there would be 2.4 million more skilled jobs in the next four years and that 44 per cent of Australia's working-age population had low literacy levels. Clearly, no one in that agency has heard of the important rule: if it looks wrong, it is wrong. And both those propositions are clearly wrong.

Getting back to the training agenda, for those of us who have been participating in this debate for too long, so much of the present plan is a clear case of deja vu all over again. The issue of providing more flexible training arrangements for apprentices has been around for decades. Why assume that there will be any real progress in the next four years? And a $100m mentoring program? Where is the evidence that this sort of thing would even work?

And we need to distinguish between those apprentices who quite rationally do not complete their training (they have learned what they need and have good employment prospects) and those from whom non-completion is a problem.

The real trouble with the government's involvement in training is that there is inadequate understanding of the fundamentals that drive investment in training. For general, transferable skills, the trainees themselves would be expected to pay in the form of lower wages as well as the direct costs.

There may be a case for the government assisting trainees with their investment, given the difficulty of securing a commercial loan backed by human capital.

But for firm-specific skills, employers should be expected to pick up a large portion of the tab as the investment is only valuable if the trainee remains at the firm. There is no case for government subsidising this form of training.

An intermediate case is that of industry-specific skills. In many instances, the larger firms undertake the bulk of the training and the small firms capture some of the benefits by poaching trained workers. But because training can involve economies of scale and scope, the only feasible option is for the training to be undertaken by these large firms.

It is understandable that some employers would be keen to palm off some of these training costs on to the government. But in doing so, other costs are incurred, including the loss of control and flexibility. This loss of flexibility is highlighted by the growing creep of regulation surrounding occupational licensing wherein the move to national licensing has actually added to the complexities.

A further heroic assumption in the training debate is that training providers are up to the task. A recent report of the Productivity Commission on the vocational education and training workforce outlines several reasons why this assumption may not hold. In particular, the report highlights the consequences of rigid industrial relations arrangements in the government-funded TAFE sector, with state-wide agreements providing for the same wages and conditions for staff in different fields. Moreover, the new Modern Award covering the sector is now hobbling private providers.

The reality is that hundreds and hundreds of millions of taxpayer money has been thrown at training through the years and a great deal has been wasted. The providers have gained but it is not clear the training has been well-suited, to the trainees or the employers. It is difficult to see this changing any time soon.

Judith Sloan is an economist and company director.