Study: Math skills at 7 predict earnings at 42

Math skills at age seven (and reading and math for girls) predicted earnings at age 42 in a British study that followed subjects from birth to middle age, reports The Atlantic.

Researchers also analyzed socioeconomic class at birth, IQ at age 11, academic motivation at 16. Controlling for other factors, “the association between basic math and reading skills and future socioeconomic status remained” and was significant.

As a next step, the researchers hope to assess the long-term impact of early education and interventions.

About Joanne


  1. Wouldn’t this go to show that one’s Math skills are innate, that you’re born with them, like a natural talent? This will give a lot of ammo to the “I’m just not good at Math” crowd…

    • It could, but based on the factors they analyzed there are several other interpretations. Maybe ages 5-7 are especially important in developing ‘math aptitude’ and good instruction at these ages is really important and has lasting benefits. Maybe kids who are good at math when they’re 7 have more perseverance than other kids. Maybe kids who are good at math when they’re young learn more science when they’re older because they have a better foundation for understanding it. Maybe being good at math gives the kids a lot of confidence, which helps them take harder classes or more risks or something else that eventually leads to higher income. These are just off-the-cuff interpretations – I’m not really advocating any of them – but there are lots of possible explanations for the data (including the one that you gave).

  2. Carrying out elementary arithmetical calculations is fairly high g-loaded.

    • What the heck is g-loaded mean?

    • palisadesk says:

      “Carrying out elementary arithmetical calculations is fairly high g-loaded”

      It’s true that there is a correlation, but high or even average intelligence (g) is not necessarily required. Some students with intellectual disabilities can do arithmetic calculations with lightning speed and accuracy. I had such a student, with an IQ of 55, who could convert fractions to decimals and percents, multiply and divide multi-digit numbers and perform operations with fractions quickly and accurately and usually without using pencil and paper.

      Performing elementary arithmetic calculations requires good rote memory, procedural and working memory, as well as practice in a variety of formats. A low cognitive ability student with these prerequisites can do arithmetic calculations well.

      *Applying* those calculations to problems, however, is a whole different story. That’s where cognitive ability plays a huge role. My low-IQ whiz kid could not solve simple word problems like “Sally has $11 and spent $5 on a movie. How much money does Sally have now?”
      He might have said $16, $55 or $2.20. Of course over time he could have been taught to discriminate which operation is used, but his ability to apply his calculating ability to problems was very limited.

      Since then I’ve taught or assessed other students with a similar profile — low cognitive ability but superior rote skills in one area or another — arithmetic, spelling and word decoding are common ones. These are called “splinter skills” when they are out of synch with the student’s ability overall.

  3. There’s also the issue of what parents do to encourage, support and teach at home, both prior to school entry and after. On a March family visit, my 6 yo first-grade grandson was happy to demonstrate his mental math skills; adding two and three-digit numbers, counting by various 2-3 digit numbers, multiplication, fractions etc. Admittedly, his school uses Singapore Math (due to parent pressure; it’s a 1-HS district), but my son and DIL also play lots of math games. On a 5″ wait in a parking lot, my son was asking J to add various combinations of parking -space numbers; 67+69, 126+145 etc, and he could do it easily, Better yet, he could explain various ways he could do such problems (60+60+16 or 70+70-4 etc.). He and his twin, and younger sister also knew which measuring cup(s) to use when we cooked together. Not all parents do that. Naturally, this ability also is significantly g-loaded.

  4. g is the general factor of intelligence. The ability to perform elementary arithmetical calculations with reasonable speed and accuracy seems to be closely related to the abilities tested by IQ tests.

  5. While I’m not a teaching profession (and thus know nothing about “g-loaded skills”), it seems to me that the ability to do math in your head is a skill highly developed by practice.

    I participated in some math competitions in high school back in the 1970s and we drilled like hell to increase our skills. The drills seemed to greatly increase our ability to do metal arithmetic and algebra. There are tons of little tricks for doing various operations in your head and learning these was very important for the competitions.

    This is not to say that mental math ability can’t also be innate, just that it’s also possible to train yourself to do it well. I’ve worked with business professionals who were unbelievable at doing business calculations in their head, because they did them all day long, every day. I used to work with a programmer who could do hexadecimal two-compliment calculations in his head. As an assembly language programmer, it was something he did a lot and thus got good at it. There is a famous passage about Feynman and Bethe that illustrates this, trying to calculate the square of 48:

    Feynman started to punch the keys anyway. “You want to know exactly?” Bethe said. “It’s twenty-three hundred and four. Don’t you know how to take squares of numbers near fifty?” He explained the trick. Fifty squared is 2,500 (no thinking needed). For numbers a few more or less than 50, the approximate square is that many hundreds more or less than 2,500. Because 48 is 2 less than 50, 48 squared is 200 less than 2,500 — thus 2,300. To make a final tiny correction to the precise answer, just take that difference again — 2 — and square it. Thus 2,304.


    • Mark Roulo says:

      Rob: “While I’m not a teaching professional (and thus know nothing about ‘g-loaded skills’), it seems to me that the ability to do math in your head is a skill highly developed by practice.”

      You know more than you think :-) g-loaded is a fancy way of saying “correlates with intelligence” or “correlates with IQ”, while trying to avoid getting side tracked with discussions about whether *THIS* particular IQ test measures the thing we are about (general intelligence, or “g”) well. Some IQ tests aren’t as good as others. This is generally considered to be a problem with the individual test rather than the concept itself.


      So … can IQ tests and math problems for 7-year-olds be prepped for? Yes, of course. If the goal was simply to increase the kids’ ability to do well on these tests we could (with time and effort) do so.


      But … and this is where a *LOT* of these sorts of studies get mis-used … the study merely shows a correlation. If you want to *predict* how groups of 7-year-olds are going to turn out socio-economically, this study shows you how to do so (with some amount of error). But the underlying assumption is that the folks being predicted don’t alter their behavior based on the study. If they do, then the correlation may well break down.


      In short, the math skills are almost certainly a proxy for something else (general smarts, parents spending time with the kids, getting to school regularly with enough rest and food, something) that does matter. Drilling our 5 and 6 year-olds on multiplication isn’t going to “help.”


      We’ve seen this before:
        *) Kids in households with lots of books do better in school. But mailing a box of books to the houses of poor folks doesn’t seem to cause the kids in those houses to do better in school (as always, with some happy exceptions).
        *) Kids who get to algebra in 8th grade do better in school (on average) than those who don’t get to algebra until 9th grade. But pushing up the curriculum to get everyone to algebra in 8th grade doesn’t seem to help.
        *) Kids who play a musical instrument do better in school, but …


      You get the idea.


      As a society, we are still looking for a silver bullet to increase kids academic performance (and maybe close the gap between rich kids and poor kids and whites/asians and everyone else). And we keep finding proxies that correlate with things. But we haven’t found a lot of things that would actually help. And those things that we have found seem to be very politicized (how to teach reading, how to teach math).

      • Right. However, spending lots of time, effort and money searching for the silver bullet (and I know of no other outfit that does this to the extent practiced by the ed world) is time, effort and money NOT spent on those things that are likely to help more kids become better educated, productive citizens; safe, orderly schools, explicit instruction in and expectations of habits and behaviors that enable success (self-control, diligence etc), good curriculum (sequential, content-rich) across all the disciplines (such as Core Knowledge, Singapore Math) and effective, efficient, explicit instruction by teachers who know their subjects. It’s not new, sexy, high-tech or any of the other buzzwords, but it works better, for more kids, than anything else we’ve tried. However, it inherently does not produce equal results, the Holy Grail of the ed world, which are impossible to achieve at any meaningful level.

  6. Using the binomial theorem like Feynman is exactly how I would have done it. All great minds think alike.

  7. It is interesting that unlike the gaps between whites and Hispanics/blacks there is little concern expressed about the gap between Asians and whites. This is so even though the gap is both substantial and increasing.