Last week, USA Today published an article, Report: More Than Half of Immigrants on Welfare. When I first saw that, I scoffed. The truth is that almost all Americans are on welfare. So I was sure that the “report” didn’t take the mortgage interest deduction into account. I was sure that they were using the standard and incorrect definition of welfare as direct aid — the definition that just so happens to make the poor look bad and the rich look good. Funny that!

But there was some push-back on the “report.” On the left, FAIR wrote, USA Today Provides a Platform for Anti-Immigration Think Tank’s Flawed Study. And on the right, the Cato Institute reported, Center for Immigration Studies (CIS) Report Exaggerates Immigrant Welfare Use. It turns out it was a typical “study” in which the conclusion was predetermined and they just played around with the data to find a way to justify it.

In particular, they looked at “immigrant-headed households.” This is a way of inflating the numbers, because children are very often US citizens. The truth of the matter is that the vast majority of welfare is not available to immigrants. Take, for example, Obamacare. But USA Today couldn’t even keep this distinction clear; it used “immigrant” and “immigrant-headed household” interchangably. So even if you accept the study as valid, the reporting by USA Today was not.

In addition to this problem, the fraction of households getting welfare does not mean much. If one immigrant-headed household has a (US citizen) child who gets free milk at school, is that equivalent to a non-immigrant-headed household that gets a couple thousand dollars per month from disability? As FAIR noted, “CIS does not provide a comparison of the cost of welfare going to immigrant and non-immigrant households — presumably because such a comparison would have weakened its anti-immigrant case.”

So what exactly is the Center for Immigration Studies (CIS)? Well, it certainly isn’t a group that objectively studies immigration. It is a fanatical anti-immigration group that puts out faux-studies for propaganda purposes. (This is distinct from Cato that does real studies, and if they don’t further the libertarian causes, they are simply buried.) In 2009, the Southern Poverty Law Center wrote, “CIS was started in 1985 by a Michigan ophthalmologist named John Tanton — a man known for his racist statements about Latinos, his decades-long flirtation with white nationalists and Holocaust deniers, and his publication of ugly racist materials.” And last year, The Daily Beast described the group as “the immigration false-fact think tank.”

But in another FAIR article, Ben Norton looked at the wider context of all this, Ignoring the Cause of Welfare: Not Laziness but Low Wages. He pointed out that the mainstream press tend to use “welfare recipient” and “freeloader” interchangeably. At one time, that might have been understandable. But there has been excessive coverage — in the mainstream press itself — of all the welfare that workers qualify for because wages are so low. Consider this, “Research conducted by the University of California at Berkeley shows that 73 percent of Americans who receive welfare are members of working families.”

It turns out that immigrants actually work more than native born Americans. What more — and I find this shocking — “American Immigration Council found that immigrants spend 45 times more in taxes than they receive in public benefits.” This is because undocumented immigrants have taxes automatically taken out of their paychecks, even though they are not allowed to benefit from most government programs. In addition, at the state level, they pay an average of 8% of their income in taxes; this is substantially higher than the 5.4% that is paid by the top 1%.

Here is the bottom line:

But it is so much easier, and so much more pleasant for the power elite, to pretend that people are poor because they won’t work. It would upset our whole way of thinking if we had to grapple with the idea that our system is unfair and that for decades, we’ve been blaming the victims.