According to your graph, GINI went up (from 34 to 38%) after taxes and transfers as well. I didn't say that it didn't change; I said it didn't change much. So

According to your graph, GINI went up (from 34 to 38%) after taxes and transfers as well. I didn't say that it didn't change; I said it didn't change much. So how do you falsify it? As with any other such case, by controlling for those variables you want to eliminate. Is it your assertion that a higher GINI somehow makes the poor (specifically poor women, if we go by that paper) less willing to study than their richer counterparts? Looking at educational attainment as opposed to enrollment is a bit, um, confusing. It was you who brought up enrollment, not I. I was speaking in general terms about tuition as a barrier to access to postsecondary education. Enrollment is a very poor proximate variable for success at college due to increasingly high dropout rates. Not that it matters, since the paper doesn't look at inequality at all. From the paper: "In this paper we document changes over time in inequality in educational attainment by family income [...]". Differences in family income, of course, mean income inequality. So unless your claim is that higher GINI makes poor kids stupid As it so happens, precisely that may be the case (we'd need further corroboration to be fully certain, but it'd hardly be surprising - stress of all kinds is known to have a deleterious effect on IQ, so stress from poverty shouldn't be any different): http://www.businessinsider.com/poverty-effect-on-intelligence-2013-8 http://www.businessinsider.com/poverty-effect-on-intelligenc... It's not something I had talked about in this thread, so this is not "my claim" in that sense, but it is another example of the pernicious effects of inequality. I give up. You have unfalsifiable beliefs which are somehow supported by academic papers on completely unrelated topics. I think you simply do not understand what "unfalsifiable" means. Practically any statement that relies on observations of natural phenomena is in principle falsifiable by making observations counter to that statement; it may just be difficult to do that. Statistical phenomena also slightly complicate the situation, though the principle remains the same. That you have a different opinion or disagree is fine - the data is hardly conclusive either way, though your interpretation does make some very counterintuitive assumptions (in particular, that people are more likely to pay for a good or service - namely, college – the less they can afford it); that you try to give this disagreement a veneer of objectivity by misusing terminology isn't so cool. Especially when you handwave away anything that doesn't fit your preconceptions.