“This is education in Liberia,” Pulitzer Prize winning American journalist Tina Rosenberg wrote in The New York Times recently.

“A girl is more likely to be married by 18 than to know how to read.”

The claim would seem at home in Liberia. The country was consumed by a civil war between 1989 and 2003. More recently its schools were closed for six months following the outbreak of the Ebola virus in 2014.

We looked at the available data.

Recent data doesn’t support claim

Rosenberg told Africa Check that her claim was based on a Liberian database from the United Nations Children’s Emergency Fund (Unicef). It estimated child marriage by age 18 at 37.9% – fractionally higher than the female youth literacy rate of 37.2%.

The most recent study of Liberia’s literacy and child marriage rate is from the country’s 2013 Demographic and Health Survey (DHS). Yet the survey’s numbers don’t line up with those shared by Rosenberg.

“I can say firmly that Demographic and Health Survey data do not support the statement in question,” the programme’s senior advisor for communication, Erica Nybro, told Africa Check.

35.9% of women aged 20 to 24 married by 18

The survey presents data in five year increments for people aged 15 to 49. Women who are 18 fall into the youngest age group of 15 to 19. But this group can’t be used to estimate child marriage as many of the woman will be younger than 18.

Instead the United Nations Children’s Emergency Fund recommends using the next age group: “Looking at the group of women 20 to 24 years old is simpler and allows for the inclusion of all girls who were married or in union by age 18 within the closest time period for which complete data are available.”

According to the 2013 Demographic and Health Survey, 35.9% of Liberian women aged 20 to 24 reported being married by the time they were 18.

The 37.9% figure cited by Rosenberg appears to have been sourced from Liberia’s previous Demographic and Health Survey in 2007.

64.2% of women aged 20 to 24 are literate

The most recent survey reported a literacy rate of 64.2% for women between 15 and 24.

The 37.2% figure Rosenberg cited is a 2007 UNESCO Institute for Statistics’ estimate for the same age group.

Someone is considered literate if they went to high school or if they can read a whole sentence or part of that sentence.

When you look at the same age group used to estimate child marriage (20 to 24) the literacy rate decreases to 58.9%.

Other organisations define literacy more strictly. For example, UNESCO doesn’t include people who can only read part of a sentence in the literate population.

If we leave out these Liberian women aged aged 20 to 24, literacy drops to 51.2% – still much higher than the marriage figure of 35.9%.

Conclusion: Author relied on old data

Rosenberg’s claim that a Liberian woman is “more likely to be married by 18 than to know how to read” is incorrect.

The 2007 data that Rosenberg based her claim on is not the most recent available.

Liberia’s 2013 Demographic and Health Survey estimated that 35.9% of Liberian women aged 20 to 24 were married by the time they were 18. In comparison, 58.9% of the same group were literate.

Researched by Kate Wilkinson, Africa Check, a non-partisan organisation which promotes accuracy in public debate and the media. Twitter @AfricaCheck and www.africacheck.org