Empirical work has shown how mathematics education exhibits certain tensions between its value as being practically useful to production and consumption on the one hand and in offering access to scarce resources on the other hand. These tensions can be ultimately traced to the contradictions in the way mathematical knowledge enhances the use value and exchange value of labour power, respectively. To understand this as a social psychological phenomenon, I look to two well-known theoretical perspectives on education, first that of the Marxist psychology of Vygotsky and activity theory (and contemporary cultural-historical activity theory) which I find tends to marginalise "exchange value". Second, I look to Bourdieu's sociology of education that tends to marginalise the use value. I then bring together these two perspectives in a joint theory of education as both development and re-production of labour power, in which use and exchange value both have their place (in commodity production). This helps explain where mathematics education might be critical.