Some serious nonsense on CBO from the WSJ

December 2nd, 2014 at 9:11 am

I almost never bother with editorials from the Wall St. Journal, which read to me something like, “Boy, if toasters could sing and dance, breakfast would be a lot more fun!”

But this AMs entry was almost like a satire—an Onion version—of their usual fare. They want to get rid of the Congressional Budget Office (and the tax revenue score-keeper, the Joint Tax Committee) because they were created, according to the Journal, by Democrats to “support the agenda of expanding government.”

In their place, the Journal’s ed board suggests that:

The wisest course is to abolish CBO and Joint Tax and allow open debate about the tax and spending implications of policy changes. Let the politicians with the authority to make these decisions be accountable for the results.

What could go wrong with that?

There’s been a fair bit of scribbling about the roles and practices of CBO since the election, with various technocratic R’s advocating that the new Congressional majority reappoint the current director, Doug Elmendorf, vs. more ideological R’s suggesting they get their own guy or gal in there.

Across the way from the singing toaster editorial, conservative writer Avik Roy has a more thoughtful piece on ways he thinks the budget agency can derive “better” numbers. (Note that he too casts a vote for Elmendorf staying on the job.) I disagree with his calls for “dynamic scoring,” which will not produce better, as in more accurate, numbers. It will just produce more numbers.

Let me explain. Dynamic scoring calls for adding macroeconomic impacts to budget estimates. If, for example, your model of the economy is jiggered to show that cutting taxes boosts productivity, labor supply, and growth, then said growth can be tapped to offset the revenue lost by the tax cut. CBO doesn’t go there, because there’s insufficient evidence that such macroeconomic feedback effects can be reliably estimated (CBO does allow for some behavioral responses to tax changes, such as timing changes in realizing capital gains).

Thus, dynamic scoring that correctly reflects this uncertainty returns a range of results—that’s what I mean by it gives you more numbers. For example, a dynamic score of Rep. Dave Camp’s tax reform plan estimated that it would raise the level of GDP over ten years by between nothing (0.1%) and a lot (1.6%). Guess which end of that spectrum fans of his plan would choose to highlight?

While I don’t always agree with CBO, I think they’re already giving us the best numbers they can given the widely accepted rules and findings of empirical economics. That’s why the director, if he or she is doing the job, doesn’t matter nearly as much as the recent dust-up would suggest. The job is not to be innovative. It’s to honestly apply the state of our knowledge.

In fact, here’s where I’d argue the budget office could improve: by being more explicit about the limits of that knowledge state. CBO’s estimates should always be delivered to the public with explicit confidence intervals, to better convey to the public the uncertainty of their guesstimates.

I was reminded of the need for this approach in their recent report on the impact of the Senate immigration bill on the economy. They reported the following: “CBO’s central estimates also show that average wages for the entire labor force would be 0.1 percent lower in 2023 and 0.5 percent higher in 2033 under the legislation than under current law.” Given the uncertainty of such an estimate—10 and 20 years away!–it seems inconceivable that we could measure such a small change with any degree of accuracy that would distinguish 0.1 percent from good old 0 percent.

In such cases, they should be very straight up about the confidence interval around their estimates, and they should do the same thing with their budget estimates (in fact, they have done so in the past; btw, dynamic scoring would make those intervals much wider). Here in Washington, we all have a bad habit of talking about estimates made under great uncertainty as if they were facts etched in stone, especially when they boost our positions.

Instead of pretending they can accurately estimate macroeconomic impacts, as Roy and the Journal would like, the CBO could help us avoid our bad habit of over-confidence in their current estimates by really hammering on the uncertainty therein.

Print Friendly

10 comments in reply to "Some serious nonsense on CBO from the WSJ"

  1. Larry Signor says:

    Promoting one ideology or another is not the job of the CBO. Reliably assessing the impact of proposed economic policy, and the costs and benefits of said policy, is their job. Yup, I can see the Congress being capable of really making a hash of that. Sort of like reviewing ones own book , movie or song. On confidence intervals; we all use some sort of internal confidence interval when appraising personal choices. I would like to see the CBO do likewise.


  2. Nick Jacobs says:

    The WSJ ed board can be peculiar, that’s for sure. Who can forget when editorial board memberDorothy Rabinowitz famously commented about the privately funded bike-share program in NYC, “Do not ask me to enter the mind of the totalitarians running this government of the city,”

    http://talkingpointsmemo.com/livewire/wsj-editorial-board-member-nyc-bike-share-program-has-begrimed-best-neighborhoods

    Great stuff!


  3. Red says:

    Toaster/Bernstein 2016


  4. bjssp says:

    Assuming that people like Douglas Holtz Eakin are somewhat on the mark about this not being a complete sham, I guess there’s something close to a legtimate argument that it might be okay to use dynamic scoring. But what are the chances that it is used properly instead of simply trying to legitimize a bunch of nonsense? Are Republicans prepared to accept estimates that suggest government spending might be a good thing–and not just defense spending? I doubt it.

    To me, this seems very fishy. I doubt there’s nothing but honest wonkery motivating Paul Ryan, et al from pressing this stuff so hard.


  5. pgl says:

    If dynamic scoring incorporated the crowding-out effects from that 1981 tax cut – it would show bigger deficits than would static scoring. Did the WSJ tell its readers this reality? Of course not.


  6. readerOfTeaLeaves says:

    Well, perhaps tomorrow the WSJ OpEd pages will advocate that thermometers have no mercury — who needs accurate data, anyway? So what if your engine melts down because the thermometer didn’t tell you the car was overheating…? Thermometers simply add to the cost of a vehicle, and you don’t need one all the time, anyway…
    The logic of these people rarely ceases to amaze me.
    ————

    To dissipate my aggravation over the WSJ braying on about the CBO, I popped on over to the Prelinger Archives and searched on ‘toasters’. And BINGO!

    The electric, pop-up toaster is used as a means to explain the significance of small-town, Americana business in “What is Business”, Coronet Educational Films (1948). Ten minutes of nostalgic delight ;-)

    Pop-up toaster at breakfast in ‘typical’ American home: 3min 50 sec
    Pop-up toaster manufacturing: 4 min 10 sec — 4 min 58 sec
    Pop-up toaster (distribution) and sales: 6 min 30 sec

    Also at minute 8, a charming, nostalgic paean to Main Street banking, a la 1948.

    Enjoy!


  7. Chuck Sheketoff says:

    Spot on, Jared. Oregon’s experience with dynamic scoring, the Oregon Tax Incidence Model (OTIM; “OH Tim”), helps prove your point. When our scorekeeper runs it on a proposal to modestly increase the state earned income tax credit, OTIM spews out results that say that the top one percent will see their income go up as a result. I can argue expanding the state EITC is good policy and even that its good for the economy, but to claim that Phil Knight will feel an impact shows dynamic scoring ultimately has no basis in reality. And when OTIM spews out job growth numbers, of course there’s no one available to figure out how to count them should the bill pass or to be held accountable to the garbage numbers.


    • Jared Bernstein says:

      Yep–it’s like when your check is wrong at the restaurant and the waitperson says, “sorry, it’s in the computer,” like that’s the end of the story. OTOH, if Tim gets the top 1% to support the state EITC…just sayin’…


  8. Dynamic scoring and budget forecasting | Bruegel says:

    […] Jared Bernstein writes that CBO’s estimates should always be delivered to the public with explicit confidence intervals, to better convey to the public the uncertainty of their guesstimates. I was reminded of the need for this approach in their recent report on the impact of the Senate immigration bill on the economy. They reported the following: “CBO’s central estimates also show that average wages for the entire labor force would be 0.1 percent lower in 2023 and 0.5 percent higher in 2033 under the legislation than under current law.” Given the uncertainty of such an estimate—10 and 20 years away!–it seems inconceivable that we could measure such a small change with any degree of accuracy that would distinguish 0.1 percent from good old 0 percent. […]


Leave a Reply

Your email address will not be published.

Current ye@r *