Meet the 'zombie stat' that just won't die
In early January, Carly Fiorina, the former CEO of Hewlett-Packard, was on ABC’s Sunday show This Week. The Senate had just confirmed Janet Yellen as the first woman to chair the Federal Reserve and by mid January, Mary Barra would take the reins at General Motors
Fiorina was asked to reflect on the moment.
"Well, clearly, there's been a lot of progress. I was the first woman to lead a Fortune 50. And as you point out, there were only seven. Now there are 23 (at) big major companies," Fiorina told host Martha Raddatz. "And you have women assuming absolute top positions of power and authority in industries and politics, as well.
"And yet the data overall hasn't shifted much," Fiorina continued. "Women remain the most underutilized resource in the world and the most subjugated people in the world. Seventy percent of the people living in abject poverty are women."
With those simple words -- "70 percent of the people living in abject poverty are women" -- Fiorina joined a line of people stretching nearly two decades to cite this powerful statistic. Hillary Clinton said it when she was first lady. Treasury Secretary Jack Lew used it twice in the past few months. Walmart included it in a 2012 report on corporate responsibility.
It would be more telling if it were true.
The deck is stacked against women in other ways, but the fact is, 70 percent of the people living in poverty are not women. Yet the statistic gets repeated and repeated -- so much that people call it the "zombie stat."
It just won’t die.
If not 70 percent, what?
Before we look at how the statistic spread, it’s important to know what the actual figure is. There is room for better data, but by the World Bank’s tally, horrific poverty does not discriminate between men and women. In a 2013 article, World Bank economists said "the poor are equally divided by gender."
The report’s lead economist Kathleen Beegle said actually, men are slightly worse off than women -- males account for 50.1 percent of the very poorest people. Beegle and her team based this on the International Income Distribution Database, a compilation of about 600 surveys across 73 countries.
There is a variant of the claim that adds girls to the equation, as in, 70 percent of the world’s poor are women and girls. Beegle said that’s wrong, too. It’s a 50/50 split through all ages, with the possible exception of the very old.
If the real mix of poverty is gender blind, where did the 70 percent figure originate? We tracked it back to the UN Development Programme. In May 1995, this UN office produced its annual Human Development Report, a massive compilation of statistics and articles to capture the condition of the people in every member nation. The report was more than 200 pages long and had one overarching theme: reducing gender inequality.
"The unfinished agenda for change is still considerable. Women still constitute 70% of the world's poor and two thirds of the world's illiterates," the report said in its foreword. The 70 percent figure gets another passing reference in the overview.
That was it, those two lines. In all the tables, indexes and pages that follow, there is nothing that looks at the gender breakdown among people who somehow survive on less than $1.25 a day, the yardstick for extreme poverty.
Who specifically authored the 70 percent statistic is a mystery.
One person who would have had a hand in crafting the two suspect sections was the director of the Human Development Report Office. In 1995, that was Inge Kaul. Kaul now is an adjunct professor at the Hertie School of Governance in Berlin, Germany. Kaul wouldn’t say if she wrote the lines that included the 70 percent statistic, but she stood by them in an interview -- for the most part.
"When you read the report, it is quite obvious, how the figure was – most likely – constructed," Kaul said "There are references to gender gaps in terms of income-earned/wages. If these gaps are around and often higher than two-thirds, then, evidently, women constitute about 70 percent of the world’s poor."
Obvious to Kaul perhaps but not to anyone else. UN statisticians don’t think those tables say that at all.
The diaspora of consultants and statisticians who were at the UN about 20 years ago recently took up the hunt for the substance behind the 70 percent. After about a month, Robert Johnston, former chief of statistical services at the UN Statistical Division, weighed in.
"I have done a little more research," Johnston said. "I conclude that the citation of 70 percent of the world’s poor are women is inexplicable except as a mistake."
Johnston and others have a hunch that this claim emerged as a mutation of partial studies in circulation in the early 1990s and before. There is mention of a 1994 report on agriculture. Some studies lumped women and children together. Drop the children from that equation, Johnston suggested, and a statistical claim is born.
A myth gathers steam
The UN report was released in May 1995. By September, First Lady Hillary Rodham Clinton repeated its words at the Fourth World Conference on Women in Beijing.
"The great challenge of this conference is to give voice to women everywhere whose experiences go unnoticed, whose words go unheard," Clinton said. "Women comprise more than half the world's population. Women are 70 percent of the world's poor, and two-thirds of those who are not taught to read and write."
Not long after Clinton spoke, the president of the World Bank, James Wolfensohn, used the 70 percent figure in a speech about the bank’s mission.
A couple of factors in the mid 1990s fueled the spread of this statistic. One had to do with the growing idea that anti-poverty efforts should focus on women. The 70 percent claim offered a clear rationale in one succinct phrase.
The other factor was pure fluke. In a case of superbly bad timing, just as the focus on women reached new heights, the agency at the UN that was all about women was going belly up.
In 1994, bad management had put the UN Development Fund for Women $20 million in the red. UN administrators launched a rescue mission. The same month that the Human Development Report came out, the Executive Board of UN Development Programme approved a strategy to get the fund back on even keel. In addition to cutting costs, the plan aimed to bring in more money from both donor countries and private foundations. To help with that, the UN invested in marketing.
The campaign continued for over a decade. We found a glossy mailer from 2006 that featured actress Nicole Kidman as Goodwill Ambassador. It asked, "Did you know ... Of the 1.3 billion people living in poverty around the world, 70% are women."
The same language lives on today. The Center for American Progress, a liberal advocacy organization, and the Thomson Reuters Foundation, the charitable arm of the worldwide news service, both posted items with the claim in 2013.
Resistance is futile
Almost as soon as this claim was born, UN statisticians began pushing back.
Alain Marcoux, a demographer at the UN Food and Agriculture Organization played out the stat to its logical conclusion. In 1997, Marcoux wrote that if 70 percent of the world's poor were women that would mean an "extra" 520 million women were concentrated in the ranks of the poorest adults. But when Marcoux plugged that figure into the estimated total number of adults in poverty, he got a strange result -- an implausible ratio of women to men.
"It does not seem that an imbalance of close to five females for each male among the adult poor has ever been observed on any significant scale," Marcoux wrote.
It’s not as though the experts had no impact. UN statisticians stopped using the number.
But the claim had spread beyond the point of containment.
One of the biggest relief organizations, CARE International, actually tried to cleanse itself of the statistic. Brian Feagans, director of communications at CARE USA, told us that in 2008, a series of memos went out across CARE’s 14 member chapters and 10,000 employees worldwide. Having concluded the statistic was wrong, they issued a policy not to use it.
Yet it showed up this year on the website of CARE’s International Directorate.
"At times, you feel like the stat enforcer," Feagans said. "A new employee comes along and they might look at the archives and see a report that came out in 2004 and they say ‘That’s great stat and it looks valid.’ And we’ll have another round of communications to warn people again."
In 2010, Duncan Green, then head of research at the anti-poverty group Oxfam GB, pleaded for advocates to stop claiming that 70 percent of the world’s poor are women. Green said all of his comrades in arms, from feminist economists to poverty researchers, agreed that the number was dodgy. Green said he was sympathetic to its persuasive power.
"But depart too far from reality ... and you undermine your own legitimacy," Green wrote, "Why should people believe anything else you say on the issue? I reckon ‘70 percent of the world’s poor are women’ crosses the line."
Green is still waiting for people to heed his cry.
"It’s the zombie stat," he said. "You kill it, and then it rises from the dead."
The formula for factoid success
Factoids are like plants, animals and bacteria. Give them the right blend of attributes, drop them in the right environment, and they will thrive.
Tom Ahern understands perfectly why the claim lives on. He runs Ahern Donor Communications and makes his living crafting phrases that evoke empathy and inspire giving.
"It has just enough of everything you want. It touches on two things people care about -- poverty and women. And the number itself makes it feel authentic," Ahern said.
In a weird way, Ahern said, the scale of the stat works in its favor.
"If you want to make a point about the poor and women, 60 percent might not be bad enough, and 50 percent would be just like men," Ahern said. "But with 70 percent, you can really make an impression."
The claim is the lucky beneficiary of something called the Goldilocks Factor. Craig Silverman is the author of the blog Regret the Error, which tracks the media’s missteps and corrections. Silverman said claims that continue to get passed from one person to the next tend to have found this sweet point of balance.
"They’re bigger than people would expect, but not so big that people find them implausible," Silverman said.
To be sure, this claim runs strongest in the community of people who care a great deal about women. Another analyst of misinformation, Brendan Nyhan at Dartmouth College, said the statistic thrives because it bolsters that underlying concern.
"You know that women are disadvantaged in many parts of the world," Nyhan said. "So this expresses some larger truth that you already believe in."
Couple these advantages with the ready availability of information on the web and in organizations’ archives and there is little doubt that if someone needs this to make their case, they will find it.
The hidden vectors of bunk
If the tale of the 70 percent claim points to anything, it is the legions of intermediaries that sit between the public and the people who actually produce original data.
Whoever wrote the foreword to that 1995 UN report was an intermediary, adding a dash of rhetorical color to an otherwise dull gray canvas of statistics and analysis. The marketers, speech writers, pundits and advocates are all doing the same thing. They represent a class of information designers.
Like interior decorators who decide where the sofa should go and the right shape for the valence, they don’t make something new. They pick and chose among the bits and pieces they can find.
When their work shows up in the words of some top official or established organization, it becomes credible. Importantly, once it’s there, there is no incentive for correction. Top officials and prominent people like Carly Fiorina who trade on their reputation don’t like to be wrong.
Treasury Secretary Lew’s press office said he would not make the claim again, but we don’t expect a correction. Repeated efforts through a variety of channels to reach Carly Fiorina produced no response. Multimillionaires hire staff to insulate them from chaff, and she may remain unaware of her mistake.
If so, she’s likely to repeat it.