Last week's budget put infrastructure spending into the spotlight. While the value of infrastructure spending cannot be disputed, sometimes more is less and the kind of infrastructure spending is more important than the amount spent.

Traffic congestion is a real problem that can be addressed by efficient policy.

Congestion is more than a tax on time; it is wastage.

In Gauteng alone, if each of the 3.1 million people aged between 15 and 65 wasted 20 minutes in traffic a day, about 1 million hours would be wasted a day, and 270 million hours would be wasted a year.

Policy to address congestion is an example of how important it is to spend wisely and illustrates that the right response is often not only obscure but counterintuitive.

Economic solutions are seldom obvious. Economic and engineering problems are very different in character. The following examples illustrate the difference between the two types of thinking.

Engineering question: How do you increase the rate of flow of water out of the bathtub?

Answer: The rate of flow of water will be increased by making the drain of the tub wider.

Economic question: How do you improve the flow of traffic into and out of metropolitan areas?

The wrong answer: One is tempted to draw an analogy between the problems and to offer the symmetrical solution to the economic question. If your answer to the economic question was to increase the number of lanes between residential and metropolitan areas, you would be dead wrong. If you had answered that increasing the number of lanes would have helped the congestion problem to an extent ... you would also have been dead wrong.

This is not merely a theoretical experiment, it is an empirical fact that increasing the number of lanes does nothing to alleviate congestion permanently. When such solutions have been adopted, congestion initially improved but within weeks it was back to the state it was before the change to the roads.

Economics is populated by counterintuitive outcomes to seemingly obvious policy remedies because we are not dealing with fluid flows but with rational economic agents who are constantly assessing the costs and benefits of their actions in ways they are often not fully conscious of.

Specifically, in the case of congestion, when people see that the lanes have widened, those who have been leaving earlier in the mornings and later in the evenings decide that it is no longer necessary to do so, car poolers decide that their transport sharing arrangements have now become redundant and some of those who were deterred from going into town for an errand during lunch will decide to take that trip. Once again you have a congestion problem.

The correct answer is a two-pronged policy response: firstly, to charge people entering commercial hubs during peak hours and, secondly, to provide sufficient and reliable public transport.

There is likely to be opposition to measures that would charge vehicles that entered metropolitan areas during peak hours.

Stores in metropolitan areas would have to pay their employees more to cover the costs of coming to work every day and would make less money if the congestion charge induced people to get their goods or services from other places. In the long run, businesses in competitive industries with price-sensitive clients might even move out of metropolitan areas.

Recently, opposition from the business community led New York to reject congestion tax.

However, we should not be defeatist. Congestion is more than a tax: the time and fuel consumed by sitting in traffic does not accrue to a central collection agency to be redistributed; it is wasted.

As South Africa ponders how best to conserve its fuel during the new era of high oil prices, it becomes more urgent to address the problem of congestion that seems to worsen in spite of the increase in the price of fuel.

Charging vehicles that enter areas prone to congestion during peak periods will be effective in alleviating congestion without prejudicing businesses in those areas only if there is sufficient and reliable public transport.

Without public transport, there will be an inelastic demand for the use of the roads and any measures to charge commuters will result in the accrual of tax revenue without much of a reduction in congestion.

A sound public transport system does not mushroom overnight but there is something we can do before such a system is up and running.

South Africa is unique in that metropolitan areas are serviced by privately owned taxis, which are important sources of income. It would be neither equitable nor wise to charge taxis for the use of congested roads. We want to encourage public transport, not tax it. Measures that exclusively target persons who travel alone into congested areas will encourage car pooling and there will be little opposition if the tax revenue accrued is earmarked for developing public transport.

How would such a system be administered? In London, cameras take pictures of car licence plates and debit owners' accounts at the end of the month. In Singapore, commuters buy cards similar to our cellular pay-as-you-go cards, which are read by scanners at checkpoints and debited accordingly.

These systems of payment have the advantage in that they are not intrusive and do not add to the congestion problem. By contrast, having traffic policemen approach vehicles with single occupants to administer fines would cause greater congestion than the single-person-occupied vehicle itself.

The most difficult matter in applying such a tax is pricing the value of using the road. Use of a road is not a good that can be bought and sold in the private sector. A price that is too high would deter business activity in excess of the cost of congestion; a price that is too low would not alleviate the congestion problem.

This is an empirical matter than cannot be settled in the course of this discussion, but a problem so serious that failure to estimate it accurately could destroy the integrity of the policy initiative.

Congestion is a real problem with real costs. It is also an opportunity to create and implement an effective policy response. Relieving congestion will result in considerable wealth and efficiency gains. It is not the most serious problem the country faces, but it is one worth taking some time to think about when you have a spare moment - perhaps when you are stuck in rush hour traffic on the way to work.