Johannesburg — IN INDIA, farmers pump water from as far as 500m underground to irrigate their crops. It's a reflection of the parlous state of India's water infrastructure, which has been allowed to run down over the past few decades. Pumping the water up from the aquifer is highly energy intensive, using something like a third of the electricity India generates.

In Australia, an area that used to grow and export rice no longer does so because of a drought more than a decade ago that wiped out the rice farmers.

These are worst-case scenarios of what can happen when the supply of water fails to meet demand. Unlike electricity, where supply and demand have to match in real time, water systems don't crash if they are in deficit. Either demand simply goes unmet in some sectors, so that some economic activities just don't happen, as has happened in Australia, or supply is created, but at extremely high (but often hidden) cost, as in India, where energy to farmers is subsidised so it's not obvious just how much that groundwater costs the economy or the environment.

A new global study conducted by McKinsey for the 2030 Water Resources Group estimates that by 2030, the world's supply of water will fall 40% short of demand. But about a third of the world's population, mostly in developing countries, will live in regions with a gap of more than 50%.

SA, perhaps surprisingly, is not that badly off, with a projected water deficit of about 17% by 2030 -- although we're drought-prone, agriculture is a relatively small part of the economy and most of it is rain-fed, not irrigated. But the 17% is an aggregate: in SA's higher growth industrial regions, the deficit is expected to be much more severe -- up to 40% in Johannesburg, for example, and 28% in Cape Town. Even the 17% aggregate could be as high as 30% if the worst fears about the effects of global warming prove true.

But what is new and useful about the Water Resources study is not that it makes alarmist noises, but rather that it attempts to quantify the scale of the potential water crisis -- and to put costs to potential solutions.

It does that in a way that aims to make the issues, and the tradeoffs, easy to understand, so that policy makers, businesses and social actors can debate them and, ideally, do something to avert the crisis. Because the surprising finding of the study, say the McKinsey researchers, is that in many countries, the water gap could be closed using existing technology, and it needn't cost a lot in the long term -- as long as countries choose an optimal mix of solutions.

These would have to include more efficient use of water, by farms, factories, cities and power stations, as well as increased supply through investment in infrastructure. They could also entail a shift in economic priorities to less water-intensive activities. And more water efficient may also mean more energy efficient: water security and energy security are often closely linked.

The methodology McKinsey uses to calculate water scarcity and develop "cost curves" for the various solutions is one that it developed for carbon emissions and was adopted at Kyoto. But unlike the global warming issue, water is a very local issue with local solutions. So the study looked at the supply and demand dynamics of 154 water basins. And it focused on four case studies, which by 2030 will collectively make up 30% of the world economy and 42% of its projected water demand: China, India, Brazil's Sao Paulo and SA. SA, say the researchers, is one of the best examples of competition from a range of users for scarce water resources.

And that is why the least-cost mix of solutions for SA will require getting everyone on board. And for SA, in particular, a mix of solutions is possible that could in fact be a net financial gain for the economy, in the sense that the returns over time would exceed the cost, if it's done right.

McKinsey talks about the various "levers" that are available and their costs and payback. Investing in infrastructure to boost supply is the most expensive way of filling the gap. SA would have to do some of that. And "supply" is also, crucially, about the quality of the water that's available and accessible, so SA would have to do a lot of that too.

But the optimal solution here would have to include a great deal of demand-side stuff. That means more "crops per drop" in agriculture . But it also means municipalities, households and industrial users need to cut leakage and introduce efficiencies of all sorts, many of which, McKinsey estimates, will pay back in less than three years. It means taking an integrated view, doing urban and energy planning, with an eye to water resources.

The report, says SABMiller 's head of sustainable development, Andy Wales, is meant to be the catalyst for a conversation about "how can we make sure we co-ordinate the way we manage water to protect economic growth".

SABMiller and another flavoured-water maker, Coca-Cola, are among the global companies that, along with the World Bank, make up the 2030 Water Resources Group and have sponsored the McKinsey study in the hope of getting that "conversation" going.

It has already been presented to the Department of Water and Environmental Affairs; it would surely be an excellent one for the new planning commission. But will the conversation go anywhere?

If electricity is anything to go by, the signs are not great. At this point, the government's record on facing up to trade-offs, taking the private sector on board, and actually making big decisions is hardly stellar. Nor is it strong on implementation. And water is so easily forgotten, as long as it keeps coming out of the tap. But it could so easily become a constraint on economic growth. Which is why it's worth taking the numbers seriously.

Joffe is senior associate editor.