CANADA READ

Yes. I think that’s a substantially stronger version of the argument than “these groups don’t deserve support.”

The key distinction you’re making is between a proxy and a cause.

Suppose we observe:

P(\text{poor outcome}\mid\text{group A}) > P(\text{poor outcome}\mid\text{group B})

That establishes a disparity. It does not establish that membership in A is the variable we should intervene on.

There could instead be some X:

A \rightarrow X \rightarrow \text{poor outcome}

or:

X \rightarrow A\text{-correlated circumstances}
X \rightarrow \text{poor outcome}

where X is something like childhood poverty, parental instability, poor schooling, neighborhood conditions, lack of collateral, lack of professional networks, discrimination in a particular decision process, etc.

If you can identify X, targeting X can be both fairer and more economically efficient than targeting the demographic correlation.

Entrepreneurship makes this especially important

Imagine two prospective founders.

Founder A comes from a demographic group whose average access to capital is worse, but personally has wealthy parents, an elite education, substantial collateral and excellent professional connections.

Founder B doesn’t belong to the targeted group, grew up extremely poor with one parent, has no family capital or useful network—but has customers, excellent unit economics and an unusually promising business.

A demographic eligibility rule can preferentially subsidize A while excluding B.

That doesn’t disprove the existence of a group-level disparity. It means the group variable isn’t necessarily the best targeting variable.

And there is something in the Canadian programs that makes your criticism more concrete. ISED says the Black Entrepreneurship Program exists specifically to address barriers faced by Black entrepreneurs and eligibility is correspondingly targeted. Futurpreneur similarly has programs accessible specifically to young Black and Indigenous entrepreneurs.

Interestingly, the government’s Futurpreneur evaluation says these tailored programs allowed people to obtain financing who “may not have been approved for the core FC loan.”

That’s precisely where I’d want much stronger evidence.

Not because those borrowers are necessarily bad investments. Rather, once you’re deliberately altering the selection threshold, the counterfactual matters enormously: