I hope so, and it’s what I wrote about yesterday in the WaPo. Let me entertain OTEers with a few more thoughts on the topic, including feedback I got yesterday.
One response was: “You’re endorsing fiscal recklessness! There’s nothing “progressive” going on here; it’s just politicians not giving a crap about our troubled fiscal accounts.”
Surely, there’s some truth to that. There are a lot of what I call deficit-chicken-hawks out there–they pretend to care about budget deficits, but as I wrote in the piece, that’s just a ruse to cut spending and shrink government. They’re happy to pass unpaid-for tax cuts all day.
But my point is that even if they’re doing the right thing for the wrong reason, the outcome will be improved. To the extent that deficits become less of an obsession, we have the potential for better fiscal policy. And I was careful in the piece to avoid going too far in the other direction. In the long term, spending must be paid for. The point is that the time for deficit reduction is when the economy is at full strength, as opposed to “always,” which is the mindless view of too many hawks, chicken or otherwise.
This raises another good question I got: why is our political system so obsessed with “payfors,” meaning there’s no proposing anything without also proposing how you’re going to pay for it? Part of this is a function of “pay-as-you-go” rules that insist Congress offset certain spending and tax cut proposals, though these rules often get waived.
As I see it, insisting on payfors for permanent tax cuts or spending increases makes sense, at least in theory (where “permanent” means something that isn’t intended to phase out, like a discretionary countercyclical program or one-time infrastructure investment). It’s in sync with that business above about outlays roughly equaling receipts over the long run.
But in the real world, the problem is immediately apparent. Since Republicans forbid any tax increases, payfors necessarily imply spending cuts. And, sticking with the real world, that means cuts to food stamps, Head Start, or social insurance, not to the mortgage interest deduction or the carried interest loophole.
In other words, insisting on payfors in this world too often implies either doing nothing (the default position) or cutting programs that poor people depend on. So, until tax revenue increases are back on the table, in the interest of fiscal equity, which admittedly isn’t the same as fiscal rectitude, I’d suggest being a lot less constrained by payfors.
Finally, based on the discussion around the figure below, a few people thought I was suggesting that CBO had a conservative thumb on the scale. Not so. They’re rigorously non-partisan.
My point was simply that their model of interest rates is generating large and persistent forecast errors, or, less technically, poorly describing how interest rates are actually moving. This has the effect of making future deficits and debt look larger than they’re likely to be (because it predicts higher costs of debt service than a lower, more realistic interest rate path would predict), and that feeds into some of the hysteria I critique.
But that’s not CBO’s fault. Their practice is to use standard modelling assumptions until the weight of the economics profession and the peer-reviewed literature suggests alterations to the standard model.
That takes a long time (though it’s notable that financial markets move a lot quicker; “forward rates”–predictions of future interest rate based on different maturities of current rates today, have significantly adjusted downwards). While the standard model undergoes needed changes–and I’ve seen a bit of evidence of progress–it’s up to the rest of us who track this sort of thing to point out persistent errors and more importantly, make noise about their implications.
In this case, that means taking scary, long-term budget projections with an even larger grain of salt than usual, and, in reference to my suspension of payfors point above, recognizing that low rates support a current investment agenda in ways that CBO may undervalue for awhile.