Pointing out that Donald Trump’s numbers don’t add up feels a little like criticizing the architectural soundness of a mud castle made by a 4-year-old. What did you expect?

Except for the niggling difference that Trump wants to become president and may ultimately have the votes to get there. So at the risk of stating the obvious, let me show you a serious problem with the Donald’s math. And, in fact, it’s not just the Donald. His problem is just an extreme version of one that’s endemic to Republican budgets, from those of Paul Ryan to the those of Trump’s competitors.

When it comes to government taxes and spending, Trump has promised the following (Marco Rubio and Ted Cruz have similar proposals):

— balance the budget in 10 years.
— not cut Social Security or Medicare.
— ramp up defense spending.
— pass a tax cut that loses $9.5 trillion of revenues over the next decade.

That can’t be done. It doesn’t matter what your hat says about America being great, your personal charisma, how rich you are, etc. As the figure below shows, you can’t get there from here. As noted, other Republicans promise similar packages, though their tax cuts are slightly less ambitious than Trump’s.

The problem: If you take that much off the table, there’s not enough left on the table to cut. To get to balance, non-interest outlays (though Trump is a serial defaulter, I’m assuming he’d pay interest on the national debt) outside of the areas mentioned would have to be cut in their entirety. Actually, more than their entirety; by 114 percent in 10 years, such that government spending outside of Social Security, Medicare, defense and interest on the debt would be -1.1 percent of GDP, as shown in the figure below.

Don’t try to figure out what that means; it’s meaningless. Perhaps it’s math like this that led to Trump’s many business failures.

What’s in the rest of government that would have to be cut? Much of the safety net (e.g., food stamps, and probably Medicaid, too, though that’s not clear from Trump’s rhetoric*) would have to go, as would the non-defense parts of discretionary spending that have already taken too big a hit, including education, cancer research and all other scientific research, road, bridges, and other infrastructure, national parks and forests, toxic waste cleanup and other environmental protection, the border patrol, the FBI, and other law enforcement, courts, homeland security including domestic anti-terrorism efforts, and veterans’ hospitals, benefits, and other services. Housing, transportation, labor, commerce — much of these functions would have to go to zero.

If you’re saying, “What’s wrong with that?,” it’s that you can’t run an $18 trillion economy with no resources devoted to infrastructure, justice, helping vets, supporting research in ways the private sector will not (it’s a “public good;” because private firms can’t count of a return on investment, they will under-invest), offsetting recessions, enforcing labor laws, negotiating with trade partners, and much else.

It’s also fiscally wrongheaded in the following sense. Suppose you actually wanted to be thoughtful about meeting the tax and spending challenges we face. [Pause here for a moment and marvel with me at how we are so intently avoiding a substantive debate about the real challenges we face…okay…we’re back live.] You’d start by asking yourself, “Which parts of the budget is the growth coming from?”

The Republican’s fiscal plans imply that the path to a sustainable budget is through cuts to the categories in the above figure. But are these the parts of the budget that are growing quickly relative to everything else?

Nope. As a new paper from some of my colleagues at the Center on Budget and Policy Priorities shows, non-interest spending on programs outside of Social Security and Medicare is already below its historical average as a percent of GDP and expected to continue to decline (see figure below).

For the record, that doesn’t imply Social Security and Medicare are “out-of-control-and-must-be-stopped!” Our aging population and health costs that have long grown faster than average prices are driving predictable and expected pressures on that part of the budget. Anyone who really wanted to make a play for longer-term budget sustainability would look there, especially at Medicare. And the smart move would be to build on the cost-control progress we see from the Affordable Care Act.

Again, there’s no reason to especially pick on Trump here. The same exercise would have Rubio cutting 90 percent of the budget on the programs in the first figure. Ryan’s budgets, but for their magical asterisks assuming the revenue lost through big tax cuts was made up through unspecified tax increases or spending cuts, face less severe versions of the same problem. Once you take Social Security, Medicare and interest payments off the table, while spending more on defense and cutting taxes deeply, all while calling for budget balance, you’re into fiscal fantasyland.

Now, we don’t need to balance the budget (we want the deficit to come down in periods of strong growth and vice versa), so one could (and should) relax that constraint. But it won’t get you out of this numbers jam. What would actually happen if we got anywhere close to tax cuts of these magnitudes would be large and growing budget deficits and debt, in both good times and bad.

So forgive me for critiquing the fiscal equivalent of a 4-year-old’s mud castle. And given that impossible things can’t occur, Trump et al.’s plans won’t be implemented. But if we ever decide to take these plans seriously, we’ll see that they are anything but serious.

* Trump has promised both not to cut Medicaid (public health insurance for the poor) and to turn it into a block grant, which would amount to a very significant cut.

Data note: The graph shows non-interest outlays as a percent of GDP but excludes spending on Medicare and Social Security benefits and annual (“discretionary”) appropriations for defense (Trump has said he will not cut Social Security or Medicare and has made comments that suggest that he will ramp up defense spending, which is assumed in the calculations).  It applies TPC’s analysis of the Trump tax cuts to current law baseline revenues, and uses a multiplier of the deficit reduction path in last year’s Republican congressional budget plan (i.e., scales it up) to produce the amount of cuts that would be needed to balance the budget in ten years.  As the graph shows, Trump would need to cut spending outside the Social Security, Medicare, and defense by 114 percent to make his budget balance, which is, of course, impossible.