It would seem to me that her question may have easily have been phrased this way, "Is marriage a greater threat to poverty than gay people?". Or, why not, "Is poverty a greater threat to gay people than to marriage?" There is as much evidence to support these claims as that made by DfU because, frankly, homosexuality is not addressed in this study.
In any case, why not conclude from studies that show marriage, and wealth, as co-factors that marriage has a positive impact on ones financial well-being? What harm can it do to pursue two worth goals instead of pitting each against the other. In any case, correlation is not causality. Poverty could, in fact, cause nothing that is measurable by this study, and marriage could have no effect whatsoever on the pocketbook.
And what about the issue of poverty? This study does appear to discuss the level of education of the participants, drawing a direct correlation between the levels of education of women, income, and marriage status of the group.
"College-educated women, whose numbers have risen sharply since 1980, often live with a partner and postpone marriage. But in most cases, they eventually marry and have children, and divorce at about half the rate of women who do not finish high school."Do any of these characteristics cause the others? Perhaps there is a correlation between those who obtain a college degree and those who escape poverty, but cause is not as easy to peg as correlation.