The big immigration news yesterday was the Heritage Foundation’s much-anticipated report claiming amnesty for illegal immigrants will cost $6.3 trillion, which the author says is a “very, very low estimate.” It’s a strange piece of white-paper-as-agitprop that other conservative groups have already condemned, as well as politicians like Jeff Flake and Haley Barbour, significantly because it undermines the rest of the institution’s push for dynamic scoring—that is, accounting for economic growth—elsewhere in the federal budget. They don’t calculate the potential revenues of legalized immigrants once they’ve started to advance in society and make more money, and fail to take into account that illegal immigrants are already a strain on the social safety net. Jim Pethokoukis and Walt Hickey are both numbers guys, so I’ll leave the rest to them.

That doesn’t mean there aren’t good reasons to kill the Rubio package, but the study has more to do with giving Senators scary-sounding factoids to cite during this week’s Judiciary Committee hearing. Andrew Stiles pulls some out:

Because the average illegal immigrant is about 34 years old, restricting access to benefits during the interim phase would have only a “marginal impact” on the long-term aggregate cost. The study found that, under current law, illegal-immigrant households produce a combined fiscal deficit of about $54.5 billion per year. After legalization, that number would fall to $43.4 billion, but would climb to $106 billion once households become eligible for welfare benefits, and would increase still further to around $160 billion during the retirement phase.

In total, current illegal immigrants would receive around $9.4 trillion in government benefits and services over the course of their lifetimes, the study found, and would pay about $3.1 trillion in taxes, resulting in a net fiscal deficit of $6.3 trillion. “This should be considered a minimum estimate,” the authors write, because it likely undercounts the total number of illegal immigrants that would ultimately be legalized and as a result would become eligible for welfare and medical benefits. “Those who claim that amnesty will not create a large fiscal burden are simply in a state of denial concerning the underlying redistributional nature of government policy in the 21st century,” the authors conclude.

That’s a bit of a dodge, since even most pro-amnesty conservatives and libertarians do recognize the welfare-state problem. Milton Friedman lays it out simply in the above video, and pretty much all sides of the immigration debate, including an author of the Heritage study, have invoked his memory. An even more succinct version is this RedState headline: “Open Borders + Welfare State = Disaster.”

In other words, these conservatives are arguing that immigration must be limited in order to preserve the welfare state, something conservatives have usually opposed. What I haven’t heard recently is the old libertarian case, at one time advanced by David Friedman and Bryan Caplan, that amnesty, legalization, and citizenship is desirable because it would undermine and perhaps destroy the welfare state. The soft version Caplan emphasizes is that immigration increases heterogeneity, which can work against the solidaristic mentality embodied by entitlement programs. Or, the hard version;  once low-earning, low-taxpaying immigrants get old enough to take advantage of Medicare and Social Security—the Heritage study’s authors concede that’s when the real costs start—they can help bankrupt them and we can all start over with something less statist.

There is denial on the pro-amnesty side, especially in the claim that amnesty would help make entitlement programs more solvent. And to those traditionalists who worry about America’s cultural continuity, to get rid of the welfare state at the cost of letting in millions of foreign-born immigrants is to give away the whole game.

Still, it’s an interesting hypothetical. When this debate resurfaced around 2008 or so libertarians tried to claim that immigrants can be legalized while being denied certain benefits, and similar ideas inform parts of the Rubio package like the “no federal benefits” provision for undocumented immigrants. While acknowledging that might be possible for large federally-administered programs, they seem to be oblivious to the complexity of the problem, with benefits disbursed at different levels of government, and with widely differing criteria for eligibility. It’s unenforceable; being selective about benefits isn’t as easy as it sounds, and in any event the requirements could easily be loosened in the face of humanitarian concerns, and there’s no guarantee, and no reason to expect, that Congress wouldn’t just change that later.