A basic tenet of the anti-globalization movement is that global economic integration is making the world a more unequal place. The rich get richer. The poor, at best, stay the same. The winners enjoy all the wonder and charm of the new technologies. The losers fall deeper into misery.

But now a new study by Professor Xavier Sala-i-Martin of Columbia University suggests that's all bunk. Measured nine different ways, the world distribution of income has become more equal, not less, over the last three decades. And although there are two billion more of us than there were in 1970, fewer of us live in absolute poverty.

Mind you, a first visit to Professor Sala-i-Martin's Web page (www.columbia.edu/cu/economics/faculty.htm) does not inspire great confidence in his results. The unsuspecting surfer is greeted by kazoo music, while a floating circle in the middle of the screen teasingly reveals shapely female thighs that turn out, once the circle expands and the full screen comes into view, to be Miss Piggy of the Muppets, dolled up in pink tights. During your visit, a pair of comic eyeballs follows your cursor around the screen. These professors are strange folk.

But rest assured, Columbia has one of the best economics departments in the world -- Canadian Nobelist Robert Mundell is on staff there, so is this year's Nobelist Joseph Stiglitz, and so was 1996 Nobelist the late William Vickrey -- and Mr. Sala-i-Martin is one of the leading scholars in his field.

Most existing studies of the world income distribution look at the distribution in individual countries and then take a world average -- so that China has the same weight as Granada. Mr. Sala-i-Martin's innovation is to use data on national income distributions to, in effect, figure out the income of just about every person in the world and then, putting everyone in the same pot, to calculate how those incomes are distributed. It's the kind of thing that's possible only with computers.

His results are a setback to the popular anti-capitalist view that every day, in every way, things are getting worse.

First, absolute poverty has declined. In 1970, 17% of the world's population made less than $1/day (in 1985 U.S. dollars, adjusted for differences in purchasing power across countries). In 1998, only 6.7% did. And although the world's population had grown by almost two-thirds since 1970, that 6.7% was actually fewer people. Similarly, where in 1970 41% of people made less than $2/day, in 1998 only 18.6% did, and that was also a smaller number of people in absolute terms -- 973 million versus 1.32 billion. Nine hundred and seventy-three million people living on less than $2/day is still far too many, but at least that number has been falling.

You get the same story no matter how you slice the data. Mr. Sala-i-Martin used the nine most popular measures of inequality that economists have come up with, and according to all nine measures, the world distribution of income has become more equal.

Unfortunately, things aren't better everywhere. The most spectacular improvements have been in Asia. The $1/day poverty rate went from 22% of Asians in 1970 to just 2% in 1998. The big success stories were China and India, which because of their size dominate the world data, but also Indonesia and Japan. China experienced rapid income growth at all income levels, but especially rapid growth at the top end, so overall inequality rose even as overall poverty fell dramatically. In Indonesia, on the other hand, $2/day poverty went from fully 69% of the population in 1970 to just 3% in 1998, while $1/day poverty was all but wiped out. And while this was happening, overall inequality actually declined, as it also did in Japan. Economic growth evidently does not always require increased inequality.

Latin America also experienced a reduction in poverty, though not so dramatic as in Asia. But in Africa things got worst. In the biggest African country, Nigeria, $1/day poverty increased by 51 million people between 1970 and 1998. On the continent as a whole, $1/day poverty increased from 22.2% of the population to 40.5% -- even as the rate in Asia was falling from 22.4% to 1.7%. Why the difference between the two regions? Asia experienced economic growth over the last three decades while Africa (with a few exceptions) did not. The effects of growth versus no-growth are staggering: "In 1970, 11% of the world's poor were in Africa and 76% in Asia. By 1998, Africa hosted 66% of the poor and Asia's share had declined to 15%."

Prof. Sala-i-Martin concludes: "If we want to reduce poverty rates in Africa, we must find a way for that continent to grow. The welfare implications of finding how to turn around the growth performance of Africa are so staggering that this has probably become the most important question in economics."

There's plenty of reason to doubt whether Jean Chrétien's Africa initiative will succeed -- golf courses, hotels and fountains won't work as well in the sub-Sahara as they did in Shawinigan -- but he is at least trying to focus the world's attention on the area of greatest human misery. And if he can help sell the message that excessive globalization isn't that continent's problem, he will have done real good.