Poor Math Skills Make a Mortgage Default More Likely

By Gemma Tarlach | June 24, 2013 2:13 pm

house mortgage for sale

Researchers investigating the trigger for the 2008 mortgage meltdown have found some borrowers’ math just doesn’t add up.

Unprecedented numbers of American subprime mortgage holders began defaulting on their loans in 2006, precipitating two years later the most severe global recession since The Great Depression. Pundits have offered numerous theories about what started the mortgage mayhem, but firm evidence has remained elusive. According to a study published today in Proceedings of the National Academy of Sciences, however, one contributor to the historic problem was poor basic math skills among borrowers.

According to researchers’ calculations, in 2006 and 2007, more than 50% of subprime mortgages in the U.S. were in default. For comparison, in 2003, about 15% of subprime mortgages were in default. Although unscrupulous lending practices in the mid-2000s have been blamed for the jump in mortgage defaults, researchers suggest that’s only part of the story. Controlling for a variety of socio-economic factors including age, ethnicity, education and household income, researchers determined that poor basic math skills correlated with likelihood to default on a mortgage.

Researchers obtained the complete loan history, including detailed payment records, of 339 subprime mortgages taken out in 2006 and 2007 in the greater New England area. During telephone interviews with the borrowers, researchers asked five questions to assess basic math skills and two questions to assess basic financial literacy. The interviewers also gave the participants a verbal fluency test designed to determine IQ. Participants’ response times to questions were also measured as indicative of general cognitive ability.

Levels of IQ and financial literacy showed no correlation with likelihood to default, but basic math skills did. Participants unable to answer questions such as

A second hand car dealer is selling a car for $6,000. This is two-thirds of what it cost new. How much did the car cost new?

were significantly more likely to be delinquent in their repayment or to be foreclosed upon.

Interestingly, participants with higher IQs were as likely to fall behind on payments as those with lower IQs, but were more likely to avert actual foreclosure, suggesting greater cognitive abilities may have helped some borrowers find a way out of delinquency.

The study’s authors stressed that their work focused on the relationship between an individual’s math skills and the likelihood of loan delinquency, rather than societal and economic factors, such as questionable lending practices, that set the stage for a sharp increase in defaults and foreclosures.

The results of their research, the authors say, suggest future default epidemics might be prevented through improved financial education for would-be borrowers and greater emphasis in school on developing math skills.

CATEGORIZED UNDER: Mind & Brain, top posts
MORE ABOUT: finances, mathematics
  • VegasLiz

    The study’s authors are dumb if they think this is thorough research.

  • genkaiharetsu

    “…greater emphasis in school on developing math skills.”

    The study seems to suggest correlation of math skills with defaulting on loans, but it doesn’t show that the former caused the latter, does it?

    • eyebeam

      Is it possible that defaulting on a mortgage causes one’s math ability to decline?

      • genkaiharetsu

        no, but both could be effects of some other factor not included in the study. just because you find a correlation doesn’t necessarily make one the cause of the other.

        • eyebeam

          What other factor do you think that could be?

          • genkaiharetsu

            i have no idea. i’m just saying that the study never definitively determined that poor math skills caused the people to default on their loans. just that there was a correlation between the two. so the suggestion that there would be less people defaulting on loans if math skills were developed is just assumption at this point. their next logical step should be to prove the cause-effect relationship, i would guess. :)

          • Metal_bender

            Genkaiharetsu is “right” – correlation does not ‘prove’ causation, as some political loudmouths are eager to point out (when it suits them). At the same time, if you can instill better math skills into a large population, perhaps any ‘lurking’ variables that _do_ cause defaults will also be influenced, and there will be fewer defaults. Establishing true causation takes a great deal of thought, logic, and sound data. The cited study establishes the correlation – a new fact. A next study can look for possible cause factors.

  • Joe Kaski

    I just finished teaching remedial summer-school math to 8th graders whom had already failed their state math test TWICE. I can tell you there is already plenty of EMPHASIS ON BASIC MATH SKILLS in our schools. But as the old saying goes, “You can lead a horse to water, but you can’t make him drink.” Despite a steady diet of real-world examples requiring basic math skills, and a steady stream of encouragement from me, many of my students refused to do much computation on their papers. “I did it in my head, Mister,” was an oft-repeated excuse for simple laziness. Sadly, this study suggests some of my lazy math students may be at an increased risk of losing a lot of money in a future foreclosure! I wish I had this story to show them last week…All the same, I hope they all passed their test the THIRD time!

    • Metal_bender

      by undertaking this group, you are earning points in Heaven! But seriously – when someone has just been told, clearly, that they messed up 2X, and many people remind them that it’s their fault for flunking what plenty of others did well, why should these ‘horses’ drink that water? Might as well be kool-aid.

      Unless you did something radically different than previous teachers of math – making the problem visual works sometimes I’ve found – I don’t hold out much hope for these kids, or for their ability to avoid foreclosures.

  • Metal_bender

    To do the listed problem, on paper I would write out the equation (translate from words to math symbols), solve for the unknown (it is a 1 variable linear equation), and report it. Making the word – symbol transition is not easy, and sometimes fiercely resisted, by most HS students. Doing it in my head, as a telephone interviewee might expect to, I have to visualize the equation in some fashion, again solve it and report. Such visualization ability is uncommon and rarely encouraged in education. I know a 3rd grader who might be able to do this problem, if his class has learned about division yet, without paper. But he is claimed to have trouble with math, since he rarely shows his work.

    In sum, I don’t think this is such a ‘simple’ math problem, and if we expect most people to be able to do such – i. e., be able to function in a mortgage-enhanced world – we will need to upgrade our expectations of HS math courses.

    Alternately, we could require each loan applicant to pass a math test before taking out a loan in their own name. I hold no hope that loan officers could be required to explain _all_ the effects of a loan clearly, and thoroughly, before giving loans to math-challenged people.

    • http://blogs.discovermagazine.com Jack Vogt

      Instead of forming and solving equations, try a direct approach: two thirds is $6000, so one third is $3000. Or try something inductive: with numbers like 6000 and thirds, the answer must be related to three. At least, something like thirteen sevenths is quite unlikely. So make a reasonable guess and then check it out.

      • Metal_bender

        if one ‘thinks’ “two-thirds is $6000,” this _is_ an equation – visualized perhaps without the written symbols, but _with_ the concepts of equality and fractions. Which the math-weak folks can’t do. And guessing by ‘something inductive’ is about equal to my 6th grade friend, who knew that two small numbers in the problem meant ‘multiply’ and one large and one small meant ‘divide.’ It got her through 5th grade, already! Perhaps we need to teach dimensional analysis explicitly – it worked for my 6th grade friend.

        • http://blogs.discovermagazine.com Jack Vogt

          Read the second sentence of your original post. I am a retired mathematician of sorts and I appreciate the power of formal methods, but sometimes excessive formalism is your worst enemy.

          I did a little thought experiment. Most of my neighbors are older and comfortably retired from successful careers and would have no trouble solving your example problem. But explain your way of doing it and you will get nothing but uncomprehending stares. You might as well be speaking Swahili.

          Re-read the second sentence. I believe you must admit your insistence upon purely (and only) formal methods has not resulted in total success.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!


See More

Collapse bottom bar