There is a null correlation between income and math anxiety, and have no explanation for the finding.

Interesting and somewhat inexplicable. From The Nature of Math Anxiety in Adults: Prevalence and Correlates by Sara Ann Hart and Colleen Marie Ganley. From the Abstract:

It is important to understand the nature of math anxiety in the general adult population, as the importance of math skills does not end when one leaves school. To this end, we present a well-powered, preregistered study of English-speaking U.S. adults describing the nature of math anxiety in this population. 1000 participants were recruited online. Math anxiety was approximately normally distributed, with the mean between “some” and “moderate”. Math anxiety was significantly negatively correlated with probability knowledge and math fluency, and significantly positively correlated with general anxiety and test anxiety. Women reported higher math anxiety than did men. Participants who had completed graduate school or had a STEM career had significantly lower levels of math anxiety than did those with less education, or non-STEM careers. Thus, we see evidence for math anxiety in U.S. adults and that it correlates with factors also reported in previous studies using younger and student populations.

From the rest of the paper, these are some of the notable findings.

A normal distribution of math anxiety in the population.

Those who are good at maths are less math anxious.

Those who are generally anxious are especially math anxious.

Women report higher math anxiety than do men, with an effect size of approximately half a standard deviation.

There is a null correlation between income and math anxiety, and have no explanation for the finding.

The gender difference in math anxiety was larger than those for math performance.

There are no race or ethnic group differences in math anxiety.

If the findings are replicated, it has some interesting implications, particularly for policy discussions around STEM gender gaps.

It is notable that for a generation, there have been significant public policy and educational initiatives to get more young women into STEM fields. Much of the focus has been around building young women's confidence as well as tackling perceived but not well substantiated biases.

This research suggests that perhaps those are misguided efforts and that the focus should be targeted on addressing anxiety, particularly given the notably large 0.5 standard deviation in anxiety between men and women.

Nothing proven, no firm conclusions, but interesting and suggestive.