Three Things You Should Know About Dynamic Scoring

The House recently changed the rules of budget scoring: The Congressional Budget Office and the Joint Committee on Taxation will now account for macroeconomic effects when estimating the budget impacts of major legislation. Here are three things you should know as we await the first official dynamic score.

1. Spending and regulations matter, not just taxes

You might think dynamic scoring is just about taxes. It’s not. Spending and regulatory policies can also move the economy. Take the Affordable Care Act. CBO estimates that the law’s insurance subsidies will reduce labor supply by 1.5 to 2.0 percent from 2017 to 2024, some 2 to 2.5 million full-time equivalent workers. If CBO and JCT do a dynamic score of the House’s latest ACA repeal, this effect will be front and center.

The same goes for immigration reform. In 2013 (and in 2006), CBO and JCT included some macroeconomic effects in their score of comprehensive immigration reform, though they did not do a fully dynamic score. Under today’s rules, reform would show an even bigger boost to the economy and more long-term deficit reduction than the agencies projected in the earlier bills.

2. Dynamic scoring isn’t new

For more than a decade, CBO and JCT have published dynamic analyses using multiple models and a range of assumptions. For example, JCT projected former House Ways & Means Committee chairman Dave Camp’s tax reform plan would boost the size of the economy (not its growth rate) by 0.1 to 1.6 percent over 2014 to 2023. The big step in dynamic scoring will be winnowing such multiple estimates into the single set of projections required for official scores.

Observers understandably worry about how the scorekeepers will do that. For example, what will JCT and CBO do with certain forward-looking models that require assumptions not just about the policy in question but also about policy decisions Congress will make in the future? If the agencies score a tax cut today, do they also have to include future tax increases or spending cuts to pay for it, even if Congress doesn’t specify them? If so, how should the agencies decide what those offsetting policies are? Does the existence of such models undermine dynamic scoring from the start?

Happily, we already have a good sense of what the agencies will do, and no, the existence of such models doesn’t hamstring them. At least twice a year, CBO and JCT construct baseline budget projections under existing law. That law often includes scheduled policy changes, most notably the (in)famous “fiscal cliff” at the end of 2012. CBO and JCT had to include the macro effects of the cliff in their budget baseline at that time, even though they had no idea whether and how Congress might offset those policies further in the future. That’s dynamic scoring in all its glory, just applied to the baseline rather than analyzing new legislation. CBO and JCT didn’t need to assume hypothetical future policies to score the fiscal cliff, and they won’t need to in scoring legislation either.

3. Dynamic scoring won’t live up to the hype, on either side

Some advocates hope that dynamic scoring will usher in a new era of tax cuts and entitlement reforms. Some opponents fear that they are right.

Reality will be more muted. Dynamic scores of tax cuts, for example, will include the pro-growth incentive effects that advocates emphasize, leading to more work and private investment. But they will also account for offsetting effects, such as higher deficits crowding out investment or people working less because their incomes rise. As previous CBO analyses have shown, the net of those effects often reveals less growth than advocates hope. Indeed, don’t be surprised if dynamic scoring sometimes shows tax cuts are more expensive than conventionally estimated; that can easily happen if pro-growth incentives aren’t large enough to offset anti-growth effects.

Detractors also worry that dynamic scoring is an invitation for JCT or CBO to cherry pick model assumptions to favor the majority’s policy goals. Doing so runs against the DNA of both organizations. Even if it didn’t, the discipline of twice-yearly budget baselines discourages cherry picking. Neither agency wants to publish rosy dynamic scenarios that are inconsistent with how they construct their budget baselines. You don’t want to forecast higher GDP when scoring a tax bill enacted in October, and have that GDP disappear in the January baseline.

I am cautiously optimistic about dynamic scoring. Done well, it can help Congress and the public better understand the fiscal effects of major policies. There are still some process issues to resolve, most notably how investments might be handled, but we should welcome the potential for better information.

For more views, see the dynamic scoring forum at TaxVox, the blog of the Tax Policy Center.

A “Normal” Budget Isn’t Really Normal

Treasury closed the financial books on fiscal 2014 last week. As my colleague Howard Gleckman noted, the top line figures all came in close to their 40-year averages. The $483 billion deficit was about 2.8 percent of gross domestic product, for example, slightly below the 3.2 percent average of the past four decades. Tax revenues clocked in at 17.5 percent of GDP, a smidgen above their 17.3 percent 40-year average. And spending was 20.3 percent, a bit below its 20.5 percent average.

Taxes, spending, and deficits thus appear to be back to “normal.” If anything, fiscal policy in 2014 was slightly tighter than the average of the past four decades.

That’s all true, as a matter of arithmetic. But should we use the past 40 years as a benchmark for normal budget policy?

It’s common to do so. The Congressional Budget Office often reports 40-year averages to help put budget figures in context. I’ve invoked 40-year averages as much as anyone.

But what has been the result of that “normal” policy? From 1975 to today, the federal debt swelled from less than 25 percent of GDP to more than 70 percent. I don’t think many people would view that as normal. Or maybe it is normal, but not in a good way.

Just before the Great Recession, the federal debt was only 35 percent of GDP. Over the previous four decades (1968 through 2007), the deficit had averaged 2.3 percent of GDP, almost a percentage point lower than today’s 40-year average.

That comparison illustrates the problem with mechanically using 40-year averages as a benchmark for normal. A few extreme years can skew the figures. In 2007, we would have said deficits around 2 percent of GDP were normal. Today, the post-Great Recession average tempts us to think of 3 percent as normal. The Great Recession has similarly skewed up average spending (from 19.9 percent to 20.5 percent) and skewed down average taxes (from 17.6 percent to 17.3 percent).

As recent years demonstrate, we don’t want a normal budget every year. When the economy is weak, it makes sense for taxes to fall and spending and deficits to rise. When the economy is strong, deficits should come down, perhaps even disappear, through a mix of higher revenues and lower spending.

Looking over the business cycle, however, it is useful to have some budget benchmarks. A mechanical calculation of 40-year averages won’t serve. Instead, we need more objective benchmarks. On Twitter, Brad Delong suggested one benchmark for deficits: the level that would keep the debt-to-GDP ratio constant. I welcome other suggestions.

Time to Fix the Budget Process

Congressional negotiators are trying to craft a budget deal by mid-December. Fareed Zakaria’s Global Public Square asked twelve experts what they hoped that deal would include. My suggestion: it’s time to fix the budget process:

Odds are slim that the budget conference will deliver anything big on substance. No grand bargain, no sweeping tax reform, no big stimulus paired with long-term budget restraint. At best, conferees might replace the next round of sequester cuts with more selective spending reductions spread over the next decade.

Those dim substantive prospects create a perfect opportunity for conferees to pivot to process. In principle, Congress ought to make prudent, considered decisions about taxes and spending programs. In reality, we’ve lurched from the fiscal cliff to a government shutdown to threats of default. We make policy in the shadow of self-imposed crises without addressing our long-run budget imbalances or near-term economic challenges. Short-term spending bills keep the government open – usually –  but make it difficult for agencies to pursue multiyear goals and do little to distinguish among more and less worthy programs. And every few years, we openly discuss default as part of the political theater surrounding the debt limit.

The budget conferees should thus publicly affirm what everyone already knows: America’s budget process is broken. They should identify the myriad flaws and commit themselves to fixing them. Everything should be on the table, including repealing or replacing the debt limit, redesigning the structure of congressional committees, and rethinking the ban on earmarks.

Conferees won’t be able to resolve those issues by their December 13 deadline. But the first step to recovery is admitting you have a problem. The budget conferees should use their moment in the spotlight to do so.

P.S. Other suggestions include investing in basic research, reforming the tax system, and slashing farm programs. For all twelve, see here.

Uncle Sam Is Smaller (Relatively) Than We Thought

At 8:30 this morning, Uncle Sam suddenly shrunk.

Federal spending fell from 21.5 percent of gross domestic product to 20.8 percent, while taxes declined from 17.5 percent to 16.9 percent.

To be clear, the government is spending and collecting just as much as it did yesterday. But we now know that the U.S. economy is bigger than we thought. GDP totaled $16.2 trillion in 2012, for example, about $560 billion larger than the Bureau of Economic Analysis previously estimated. That 3.6 percent boost reflects the Bureau’s new accounting system, which now treats research and development and artistic creation as investments rather than immediate expenses.

In the days and months ahead, analysts will sort through these and other revisions (which stretch back to 1929) to see how they change our understanding of America’s economic history. But one effect is already clear: the federal budget is smaller, relative to the economy, than previously thought.

Picture1

The public debt, for example, was on track to hit 75 percent of GDP at year’s end; that figure is now 72.5 percent. Taxes had averaged about 18 percent of GDP over the past four decades; now that figure is about 17.5 percent. Average spending similarly got marked down from 21 percent of GDP to about 20.5 percent.

These changes have no direct practical effect—federal programs and tax collections are percolating along just as before. But they will change how we talk about the federal budget.

Measured against an economy that is bigger than we thought, Uncle Sam now appears slightly smaller. Wonks need to update their budget talking points accordingly.

Immigration, Dynamic Scoring, and CBO

Immigration policy poses an unusual challenge for the Congressional Budget Office and the Joint Committee on Taxation. If Congress allows more people into the United States, our population, labor force, and economy will all get bigger. But CBO and JCT usually hold employment, gross domestic product (GDP), and other macroeconomic variables constant when making their budget estimates. In Beltway jargon, CBO and JCT don’t do macro-dynamic scoring.

That non-dynamic approach works well for most legislation CBO and JCT consider, with occasional concerns when large tax or spending proposals might have material macroeconomic impacts.

That approach makes no sense, however, for immigration reforms that would directly increase the population and labor force. Consider, for example, an immigration policy that would boost the U.S. population by 8 million over ten years and add 3.5 million new workers. If CBO and JCT tried to hold population constant in their estimates, they’d have to assume that 8 million existing residents would leave to make room for the newcomers. That makes no sense. If they allowed the population to rise, but kept employment constant, they’d have to assume a 3.5 million increase in unemployment. That makes no sense. And if they allowed employment to expand, but kept GDP constant, they’d have to assume a sharp drop in U.S. productivity and wages. That makes no sense.

Because increased immigration has such a direct economic effect, the only logical thing to do is explicitly score the budget impacts of increased population and employment. And that’s exactly what CBO and JCT intend to do. In a letter to House Budget Committee Chairman Paul Ryan on Thursday, CBO Director Doug Elmendorf explained that the two agencies would follow the same approach they used back in 2006, the last time Congress considered (but did not pass) major immigration reforms.

In scoring the 2006 legislation, JCT estimated how higher employment would boost total wages and thus increase income and payroll taxes, and CBO estimated how a bigger population would boost spending on programs like Medicaid, Food Stamps, and Social Security. They found that the legislation would boost revenues by $66 billion over the 2007-2016 budget window and would boost mandatory spending by $54 billion; various provisions also authorized another $25 billion in discretionary spending subject to future appropriations decisions.

I remember that estimate well since I was then CBO’s acting director. At the time, I thought this was a pretty big deal, doing a dynamic score of a major piece of legislation. I expected some reaction or controversy. Instead, we got crickets. It just wasn’t a big deal. The direct economic effects of expanded immigration—bigger population, bigger work force, more wages—were so straightforward that folks accepted this exception from standard protocol. I hope the same is true this time around.

Note: The approach CBO and JCT will use in scoring immigration legislation is only partially dynamic. It accounts for the direct effects of increased immigration, such as a bigger population and labor force, but not indirect effects such as changing investment. In other words, it follows the standard convention of excluding indirect changes in the macroeconomy; the innovation is accounting for the direct effects. We used the same approach in 2006, analyzing indirect effects in a companion report separate from the official budget score. CBO and JCT will take the same approach this time around.

The Balanced Budget Amendment’s $300 Billion Error

The balanced budget amendment introduced by Senate Republicans yesterday contains a striking error. As written, it would limit federal spending much more than they claim or, I suspect, intend (I said the same back in 2011, when this first came up).

The senators want to balance the budget by limiting spending rather than raising tax revenues. They thus propose the following, according to a press release from sponsor Senator John Cornyn:

Requirement to Balance the Budget. With limited exceptions, the federal budget must be balanced.

Presidential Requirement to Submit a Balanced Budget. Prior to each fiscal year, the President must submit to Congress a balanced budget that limits outlays to 18 percent of GDP.

18 Percent Spending Cap. With limited exceptions, Congress must limit outlays to 18 percent of GDP.

That 18 percent figure is in line with average tax revenues over the past four decades, but well below average spending, which has been about 21 percent.

So what’s the error? The way the amendment would implement the spending limit:

Total outlays for any fiscal year shall not exceed 18 percent of the gross domestic product of the United States for the calendar year ending before the beginning of such fiscal year, unless two-thirds of the duly chosen and sworn Members of each House of Congress shall provide by law for a specific amount in excess of such 18 percent by a roll call vote. (Emphasis added.)

The amendment thus doesn’t limit spending to 18 percent of the current fiscal year’s GDP; it limits it to 18 percent of GDP in the previous calendar year.

At first glance that may not sound like much. But it works out to be 21 months during which inflation and real growth will almost always be boosting GDP. For example, fiscal 2014 starts in October of this year. If the amendment were effective today, spending would be limited to 18 percent of last year’s GDP—that’s calendar 2012, which started (of course) in January 2012.

That 21-month lag has a big effect on the spending limit. Consider fiscal 2018, the first year it could conceivably take effect (because of a waiting period in the amendment). The Congressional Budget Office projects that nominal GDP that year will be $20.9 trillion. So the Republicans’ fiscal 2018 spending limit ought to be 18 percent of that, a bit less than $3.8 trillion. But the amendment would look back to calendar 2016 to set the limit. CBO estimates that year’s GDP at roughly $19.1 trillion, nearly $2 trillion less than for fiscal 2018. The amendment would thus limit fiscal 2018 spending to a bit more than $3.4 trillion. That’s only 16.4 percent of GDP that year, about $330 billion less than the Republicans’ stated goal.

If you do the same math for the remaining years in CBO’s latest outlook, fiscal 2019 through 2023, that gap never falls below $300 billion.

The same drafting error came up when GOP senators introduced a balanced budget amendment in 2011. When I wrote about it then, several commentators suggested that perhaps it wasn’t an error, but rather a sneaky way to try to limit spending even further. I am not so cynical. Drafting a spending target based on GDP isn’t easy, since you don’t know what future GDP will be. So I can understand why someone drafting this might try to use a measure of GDP that’s already known, albeit subject to much revision. But they goofed.

It’s disappointing that no one has fixed this error in the intervening 18 months. I am not a fan of an arbitrary constitutional limit on spending—even with a supermajority escape valve—but as a fan of arithmetic, let me offer one simple approach: use a GDP forecast from whatever entity is responsible for the spending forecast. For the president’s budget submission, that would be the Office of Management and Budget, and for the congressional process it would be either CBO or the House and Senate Budget Committees. That would make the GDP forecast even more politically sensitive, of course, but it’s better than a formula that misses its intended target by $300 billion each year.