OMB Director Shaun Donovan criticized House Republicans today for their rule change on dynamic scoring. I’ll skip the preliminaries and offer responses to his main critiques.
OMB Director Shaun Donovan: “Using dynamic scoring for official cost estimates would risk injecting bias into a broadly accepted, non-partisan scoring process that has has existed for decades.”
No, the new rule attempts to eliminate an oversimplifying assumption that we know is incorrect and biased in a limited number of cases where that bias is large enough to matter. We know that some policy changes can increase (or reduce) the size of the economy, and that to assume otherwise is wrong. The longstanding scoring process is biased against policies that would increase economic growth, and biased for policies that would shrink the economy. The size of the effect of large and broad-based reductions in tax rates is uncertain, but we’re pretty darn sure it’s not zero. Certain immigration reforms would increase domestic labor supply and increase economic growth. More accurate scoring would incorporate both types of effects.
Donovan: “[D]ynamic scoring requires CBO and JCT to make assumptions in areas with unusually great uncertainty.”
- Yes, these assumptions are uncertain. So were CBO’s estimates of short-term fiscal multipliers from the 2009 stimulus law. CBO assumed (guessed) that the effect on short-term GDP of increasing government of purchasing goods and services by $1 could range from 50 cents of increased GDP on the low end, to $2.50 of increased GDP on the high end. (See Table 2 here and this paper.) That’s a big range that didn’t dissuade Team Obama from arguing that fiscal stimulus would increase GDP growth.
- Is it better to be precisely wrong or imprecisely right? I’m generally for the latter, which is what this rule change would encourage. We know that assuming zero GDP effect from large fiscal policy changes is incorrect. I think it’s an improvement to move to a new non-zero point estimate even if that means we have to accept wider error bounds.
- The way you get better at narrowing these uncertainties is to do more estimates and to learn by doing.
Donovan: “Dynamic scoring would require CBO and JCT to make assumptions about policies that go beyond the scope of the legislation itself.”
- Yes, just like they had to do on fiscal stimulus in early 2009. They had to make an assumption then about how the $800+ B of increased debt would be financed in the future. They also had to make an assumption about whether the Fed would turn the monetary policy dial differently after Congress had pulled hard on the macro fiscal policy lever. Nothing new here, they’ll have to do for a few more bills what they have been doing elsewhere for quite a while.
Donovan: “Dynamic scoring can create a bias favoring tax cuts over investments in infrastructure, education, and other priorities. … The House rule would not apply to discretionary spending, ignoring potential growth effects of investments in research, education, and infrastructure.”
- So you’re opposed to dynamic scoring for policies you don’t like, but OK with it for policies you do like? Hmmm…
- This is almost a conceptually valid critique but it’s a little hard to see how one would practically expand the new rule to apply to discretionary spending. The House rule would only apply dynamic scoring to big fiscal policy changes, those with an aggregate static budget impact bigger than 0.25% of GDP (>$45 B per year in 2015). That high de minimis threshold is a smart practical limitation so that CBO doesn’t have to worry about the growth impacts of the cars-for-crushers program or a $50M increase in spending for pet project X. But since nobody is proposing adding >$45B/year to discretionary spending, this limitation in the new rule has no practical impact.
- But the House rule doesn’t create a bias for tax cuts. It eliminates a pre-existing bias against very large policy changes that would expand the supply side of the economy, including but not limited to broad-based reductions in tax rates.
The Obama Administration’s position seems to be:
- We like incorporating the short-term growth effects of increased government spending (fiscal stimulus);
- In that case we’re OK with estimators making assumptions that go beyond the scope of the legislation (Fed reaction to fiscal stimulus);
- We’re OK with lots of estimating uncertainty when it applies to policies we like (big ranges for short-term fiscal multipliers);
- We’re happy to trumpet the growth benefits, as estimated by CBO, for policies we like such as the Senate immigration reform bill;
- But we don’t want to incorporate long-term supply-side effects from policies we don’t like (broad-based tax relief),
- And we will label any scoring change which incorporates (but is not limited to) those policies as “biased.”
- Their goal is to preserve a longstanding scoring bias that keeps taxes probably higher than they otherwise would be if scoring were more accurate and policy neutral.
In contrast, my view is:
- Scoring should be accurate, unbiased, and policy neutral.
- The “fixed nominal GDP” estimating convention is both inaccurate and biased.
- Estimates of very large fiscal policy changes should incorporate CBO and JCT’s best judgment (not Members’ of Congress best judgment) about both the short-term demand-side GDP effects and the long-term supply-side GDP effects.
- Relaxing this fixed nominal GDP assumption should apply to all types of fiscal policy changes exceeding a certain size, whether I like them or not. The House’s >0.25% of GDP seems like a reasonable threshold.
- CBO and JCT should explain their estimating ranges and uncertainty for these dynamic estimates, as they have for short-term fiscal multipliers in the past.
- The House rule is a responsible improvement in scoring policies that makes scoring more accurate by reducing a long-standing bias.The increased uncertainty that results is worth reducing the underlying bias.