Dynamic scoring is a good idea with big problems

Congressional Republicans are set to appoint economist Keith Hall as the new director of the Congressional Budget Office. But more importantly, they’re going to ask Hall to change the way the CBO scores big bills. They want to get dynamic.

I’ve spent a fair amount of time over the last few weeks trying to figure out whether it’s a good idea for the Congressional Budget Office and the Joint Committee on Taxation to use dynamic scoring. The answer, as best I can tell, is that dynamic scoring is a great idea — if congressional scorekeepers can pull it off. But they probably can’t pull it off. And doing dynamic scoring poorly is worse than not doing it at all.

The idea behind dynamic scoring is pretty simple: it just means taking the broader economic effects of policies into account when trying to figure out how much they’ll cost.

The canonical example is one that conservatives like: tax cuts, if properly designed, can juice economic growth. More economic growth means more revenue. More revenue offsets part of the tax cut’s cost. A static score — one that didn’t take economic effects into account — would miss that.

But there are examples liberals like, too. For instance, during the recession, stimulus spending boosted economic growth. A dynamic score of the cost of the stimulus would have shown it to be quite a bit cheaper than the static score suggested. Economists Brad DeLong and Larry Summers estimated that it may actually pay for itself.

This gets to a core point about dynamic scoring that liberal critics often underplay: the world is dynamic, and while there are all kinds of ways in which dynamic scoring can go wrong, we know, for sure, that static scoring is wrong.

But dynamic scoring is going to be wrong too. The question is whether dynamic scoring, in practice, is any less wrong than static scoring.

The problem with dynamic scoring is also pretty simple: it’s really, really hard to do. And not just because it’s technically difficult, though it is. A lot of the judgments that need to be made to produce a dynamic score are judgments about the future that the Congressional Budget Office and the Joint Committee on Taxation are in no position to make.

Here’s an example. There’s a fair bit of research that suggests well-designed tax cuts can grow the economy — if they’re paid for with spending cuts. But if those tax cuts are paid for with future tax increases, they don’t help the economy. So imagine you want to create a dynamic score for tax cuts that Congress is, for now, charging to the national credit card. How do you do it?

The best answer is probably that you don’t do it. Sometimes the only way to win is not to play. But Congress isn’t going to accept a shrug. Instead, the CBO and JCT will have to start making a bunch of assumptions they have no real business making. Harvard’s Greg Mankiw, who served as chief economist to President George W. Bush, runs through some of them:

Nor is it just Congress whose behavior the CBO will now have to predict. It’s also the Federal Reserve. If Congress passed a massive deficit-reduction bill and, in response, the Federal Reserve decided to lower interest rates a bit, that would have a huge effect on the cost of the bill under dynamic scoring. But how is the CBO supposed to read Janet Yellen’s mind three years from now?

Then there’s the more prosaic, but insidious, doubts that dynamic scoring lets into the process. By adding complexity, it adds more opportunities for political manipulation and partisan skepticism.

The CBO isn’t particularly transparent in how its models work now. But in a dynamic scoring world, the assumptions built into those models become much more important — and much more contestable. Which study the CBO and JCT end up using on the elasticity of labor in the face of higher marginal tax rates can decide whether a tax hike looks like a great idea or a disaster.

If you want to get really sinister about it, as the models require more assumptions, there’s more opportunity for tampering. But you don’t even need to go that far. More complex, assumption-dependent models increase the importance of sincere disagreements over which research to believe.

Alan Viard, an economist at the conservative American Enterprise Institute, doesn’t worry over this too much. “Many critics of dynamic scoring say they oppose it because it would increase the potential for manipulation,” he says. “I don’t know that’s a very good argument against it. Someone who wanted to deliberately manipulate the score has so many mechanisms to do it now that I doubt keeping this one mechanism out of their hands will be an effective constraint.”

I think Viard’s rejoinder is right, but it underscores the problem. The trust the CBO and JCT have won is a rare and fragile thing. There’s a constant hum of criticism around the work that they do. There are fears, even now, that the models are wrong, the work is politicized, the results are garbage. But CBO and JCT are protected, in part, by being pretty straightforward.

Dynamic scoring will make those models a lot more complex, and it will make the results a lot more uncertain. Moreover, the very idea of using dynamic scoring is politically controversial, and since even believers in dynamic scoring recognize the huge errors that can lurk in its equations, the scores it returns are going to command that much less authority.

All that said, I’m not actually opposed to seeing CBO and JCT begin experimenting with dynamic scoring. If carefully explained dynamic estimates were attached as supplementary material to scores of big bills, I think that would be worth doing (and, indeed, CBO and JCT did exactly that when they scored the Senate immigration bill). But forcing CBO and JCT to produce definitive dynamic scores for big bills, as congressional Republicans are doing, is a bad idea — and one Republicans might come to regret.

Perhaps the simplest way to put this is that Republicans with dreams of cheaper tax cuts dancing in their heads should ask themselves how they would have reacted to dynamic scores that found the stimulus bill would have cost far less than static estimates showed. That’s how Democrats are going to react to the scores for their tax cuts. And, if there’s a President Hillary Clinton, that’s how Republicans are going to react to the dynamic scores for universal pre-k. Is that really going to improve Washington’s policy debates?