Johannesburg — WITH all this debate about whether we need a tax on "hot money" inflows, no one has asked why we're treating the symptoms, as it were, and not the cause.

The problem is SA has become heavily dependent on inflows of foreign capital, because it invests a lot more than it saves (and imports more than it exports). Those inflows tend to make the rand volatile, because the money can leave as suddenly as it arrives; they have also in recent years made the rand stronger, even though SA's economic fundamentals suggest it should be weaker.

One solution that has been proposed, most recently in an African National Congress discussion document, is a "Tobin tax" on speculative inflows, similar to that imposed by Brazil and Chile. The thinking is that this would help to deter those flows and so help to weaken the rand. Whether it would is not at all clear (Brazil's worked only briefly) . It's even less clear how one would define "speculative" flows as opposed to others, and where one would draw the line between desirable and undesirable investments by foreigners. Indeed, some of the flows that have driven the rand up most sharply in recent years have been related to big foreign direct investment deals, not "hot money" at all - and no one seems to be suggesting it would be a good idea to deter those deals.

Nor does anyone seem to be suggesting that if we wanted to be less dependent on foreign capital, we could always start by saving more. That is in any event what we should be doing as a country if we want to sustain faster rates of economic growth.

At a theoretical level, it is not very clear why higher savings rates should be associated with higher investment and growth rates - why it should matter, for example, whether domestic investment is financed by domestic savings or by foreign savings. But all the empirical evidence is that high-growth countries have high savings rates. China's gross savings rate is an extraordinary 53% of gross domestic product. India's savings rate is about 34%. SA's is 16%. That is at least better than the US's 12%.

But when it comes to household savings, we are up there with the most conspicuous of the consumer nations: at a mere 1,6%, SA's household savings rate is one of the worst in the world. What savings SA does have are largely corporate; the government was a net saver for a while during the boom years but it is now back to dissaving. Of course it is in part a culture thing. Not for us the prudent habits of Chinese households: given a choice, we spend it rather than save it every time.

But it's not just that we lack a savings culture. There are some structural reasons for SA's low savings rate.

One is the demographic shift, with the expansion of the middle classes and new emerging middle-class consumers who haven't previously owned houses or cars and have been acquiring these over the past decade, rather than necessarily saving for their old age or their kids' education. That structural shift may be coming to an end.

But a more fundamental structural factor is SA's very low employment rate. Only one in four adult South Africans has a job, so dependency ratios are high, with each income supporting many mouths. That affects low- income households particularly, limiting their ability to save, but it's also a factor among the middle classes. So creating jobs is probably the most important thing SA could do to boost household savings.

Ironically, perhaps, SA's fairly extensive social safety net also tends to put the lid on savings behaviour among low-income earners. It's not just that people may feel they don't have to save for retirement because they can rely on the state's old-age pension - there is actually a disincentive to save for retirement if you are a low-income earner because if you do accumulate savings you may fail the means test and lose your entitlement to the state pension. So abolishing the means test could help. So, too, could a social security system to which all employed people are required to contribute.

New bank liquidity rules that are soon to be introduced internationally could also have an effect because they will give the banks an incentive to go all out to attract retail deposits and therefore to make it more attractive to put your money in the bank instead of spending it. But though it may be easy to see why the country needs to save, it may be harder to persuade individual households, especially low-income households, of the merits of saving. One thing that may get them thinking is new consumer protection legislation that forces retailers and others to disclose exactly what it costs to buy on credit not cash.

But though there are good social and economic reasons for households to save, at macro level it depends who is doing the saving. Even if many thousands of low income earners upped their savings rates, it might not have a huge effect on SA's savings rate because the absolute amounts of money would be quite small. But if even a small number of very affluent households were to save a much higher percentage of their incomes, the totals could be quite large. For that, however, SA would have to look at measures such as tax concessions to give higher income earners incentives to save. But that's not up for debate.

Joffe is senior associate editor.