The phrase "tipping point" has become one of the cliches of our time. It covers not only the edge beyond which there will be (alleged) catastrophe but also mundane matters of human interaction. Moreover, as the 2008-09 budget indicates, the tipping can occur in either direction.

Take the figure of $150,000 a year. On page four of Wayne Swan's budget speech, $150,000 a year for couples is the tipping point below which they will no longer have to pay the Medicare levy surcharge. Swan says this is not a high family income and the increase in the income threshold from $100,000 to $150,000 will make the levy "fairer".

Now turn to page six. Here $150,000 a year for couples is the tipping point above which they will no longer be eligible to receive the $5000 baby bonus. Swan says that $150,000 is a high family income and, consequently, this payment should be subject to a means test.

This distinction does not make sense. It seems that the decision is best explained by a two-dislikes analysis. Some members of the Labor Government do not like private health insurance and are opposed to a surcharge that encourages taxpayers to take out such cover. Also, some Labor MPs do not like the previous government's baby bonus policy and favour restricting it in a way that minimises a political reaction.

Yet the baby bonus was an effective form of maternity payment. First, it was paid by the government and, consequently, did not impose any financial burden on employers. Second, it did not discriminate against women who were not in the workforce. Third, women on high, medium, low or zero incomes all received the same amount.

Under the Howard government scheme, a woman on an income of $200,000 a year would receive $5000 - as would a woman on $30,000 a year. There is no doubt as to whom the baby bonus was more significant. Yet the higher income earner received a payment that would financially assist with the birth of her baby.

By formally means-testing the baby bonus, Kevin Rudd has effectively abandoned a form of maternity leave for couples who are regarded as well off. This in spite of the fact that Rudd said on April 29 last year that a $250,000 family income is not a lot of money in expensive parts of Australia.

However, the Rudd Government has directed the Productivity Commission to inquire into paid maternity, paternity and parental leave and to submit a final report by the end of February. Public hearings have just concluded and the commission will release a draft report in September.

The problems facing the Productivity Commission and, in time, the Rudd Government were evident in the submission which the commission received last week from the Australian Retailers Association. It argued that any paid maternity scheme must be fully funded by the Government and that small retailers could not afford to match the difference between a government payment, equivalent to the minimum wage rate, and the rate of pay which the employee was receiving before she ceased work for her pregnancy.

The association maintained that if small retailers had to assume any financial responsibilities for maternity leave they would hire men or older women and/or resort to employing casuals.

In this context, size does matter. Already many large businesses have paid maternity schemes. This even applies in the retail sector, where Aldi, Myer and Woolworths have introduced such provisions. Also, generous maternity provisions already apply in the Commonwealth, state and territory public services along with local government and the tertiary education sector.

The policy makes good economic sense, since it helps an employer to retain valuable staff, because it can be readily funded. Small business, however, is quite another matter.

It is interesting to note that many of the high-profile advocates of maternity leave have never run a small business and do not appear to understand that, for such enterprises, both paid and unpaid leave poses considerable difficulties. Many advocates work, at the taxpayers' expense, in such bodes as the Human Rights and Equal Opportunity Commission and the Australian Institute of Family Studies. Others are employed in employee organisations like the Australian Council of Trade Unions.

Some do not understand just how difficult it is for small businesses to make up salary packages each week or to cover for an employee who has an entitlement to, say, 52 weeks' unpaid leave with a right to return to a job of equal value.

Maternity leave works best when it is not regulated. If the Commonwealth Government wants to pay maternity and paternity leave to all new parents this is fine - provided, of course, such a policy can be funded. Likewise, if businesses want to provide maternity/paternity entitlements as part of an employment package - that's also fine.

The problem emerges when bodies like HREOC and the ACTU want business to supplement a government-funded scheme, irrespective of their economic circumstances, or when they make unworkable demands of a non-wage kind on small businesses.

Maternity leave regulation will prove a disincentive to small businesses employing young females. Yet, left alone, these employers (quite a few of whom are women) can usually work it out with their female workforce.

Many small businesses do not mind mothers bringing very young or sick or holidaying children to work. Such practices are much more difficult in medium to large enterprises. Also, small business can readily negotiate unpaid leave if not burdened by the threat of legal encroachment for breach of some regulation. The Productivity Commission would be well advised not to interfere with such existing informal arrangements. There's a tip.

Gerard Henderson is the executive director of The Sydney Institute.