His point is about married women providing marginal revenue to a household. Household upkeep has real costs (cooking, cleaning, taking care of the kids, etc) and a housewife is a cheap way to provide the labor for it. High marginal tax rates (when couples are taxed as a single unit — Canada taxes couples individually) can easily tip the cost/benefit from working to staying at home. As you note, removing these women from the labor pool this has real effects on GDP.

In the married women case, child care expenses are the real driver. In the US, we make these tax deductible. This help make work “affordable” but costs are very high.

In my unscientific sample of 1 (me), the wife would love to work. However the (Expected Pay – child care) * (1 – marginal tax rate + payroll taxes) – (Eating out More – Other Incidentals) is close enough to being 0 that it is not worth the headache.

If we are looking for a policy that puts more people into the labor market, then a tax scheme like Canada where the wife’s income is taxed separately could make a lot of sense. Dropping the tax rate from 28 (or 33 or whatever) to something in the low teens would put a lot of women in the workforce, generate higher GDP, generate more tax general revenues, and generate greater payroll tax revenues.