The Center's work on 'Process' Issues

New York Times Warns Against “Dynamic Scoring”

December 8, 2014 at 12:06 pm

A New York Times editorial this weekend raised several red flags about so-called “dynamic scoring” — that is, including estimates of the macroeconomic effects of policy changes in official cost estimates for tax and spending legislation.  We strongly agree.  Our recent paper making the case against dynamic scoring, and a short summary we released today, explain that:

  • Current budget estimates aren’t “static.” The Congressional Budget Office (CBO) and the Joint Committee on Taxation (JCT) incorporate in their cost estimates many changes in individuals’ and companies’ behavior in response to proposed changes in tax rates and other policies.
  • Dynamic estimates are highly uncertain. Different models and assumptions produce widely varying estimates of how policy changes would affect the overall economy.  Some models’ results depend on assumptions about how future Congresses will reduce deficits.  And the models all have significant gaps.
  • Dynamic estimates are prone to manipulation. Because of this uncertainty, congressional leaders will likely cherry-pick the model and assumptions that give the most favorable estimates.  That’s exactly what House Ways and Means Chairman Dave Camp did in touting the highest estimates of economic and revenue growth for his tax reform proposal — estimates more than ten times greater than JCT’s lowest ones.  (See figure.)
  • CBO did not use dynamic scoring for the 2013 Senate immigration bill. Some members of Congress claim, incorrectly, that CBO used dynamic scoring to estimate the bill’s budgetary effects.  CBO’s official cost estimate took account of the bill’s direct effect on the U.S. population and labor force.  But it did not include estimates of the bill’s more speculative and uncertain effects on the economy, such as its effects on investment and productivity.

You can follow me on Twitter at @PaulNVandeWater and my co-author Chye-Ching Huang at @dashching.

“Dynamic” Estimates Are Highly Uncertain, Subject to Manipulation

November 17, 2014 at 5:10 pm

An American Action Forum event today to promote “dynamic scoring” for tax and spending legislation unintentionally illustrates what Chye-Ching Huang and I explain in a newly updated paper:  estimates of the macroeconomic effects of policy changes — which is what dynamic scoring would include — are highly uncertain and subject to manipulation, so they shouldn’t be part of official cost estimates.

In reasonably balanced remarks, Senator Orrin Hatch (R-UT) said that “we should not expect dynamic scoring to produce outsized miracles from either the supply side or the demand side.”

But Tax Foundation President Scott Hodge, in giving his organization’s estimates of the effects of several tax proposals, promised just such miracles.  According to Hodge, cutting the corporate income tax rate or allowing full expensing of investments (that is, allowing firms to deduct the investments’ full cost from their taxable income up front, rather than depreciating it over the investments’ lifetime) would more than pay for itself by boosting economic growth and, in turn, tax revenues.

That’s highly implausible.  But it shows how advocates can manipulate assumptions or cherry-pick dynamic-scoring estimates to buttress their agenda.  Ways and Means Committee Chairman Dave Camp (R-MI) did the same thing when he cited only the most optimistic of many “dynamic” estimates in touting the benefits of his tax reform proposal, as our paper and the graph below show.

A Dangerous Way to “Fix” American Government

October 21, 2014 at 4:21 pm

“A dangerous proposal is circulating in states across the country that could widen political divisions and jeopardize cherished rights and freedoms,” CBPP President Robert Greenstein explains today in the Washington Post’s PostEverything blog.  He continues:

The push is coming primarily from well-organized, arch-conservative groups seeking to capitalize on the decline in public trust in government to limit the federal government’s role and spending powers.  And the method they prefer is a constitutional convention — the first since the 1787 conclave that produced the U.S. Constitution.

Under the Constitution, if two-thirds of state legislatures call for a convention to amend it, one must be convened.  Some of those pushing for a convention say that 24 of the needed 34 legislatures have approved such resolutions.  Advocates of a convention have targeted more than a dozen other states and are developing lobbying campaigns to push for such resolutions there.

The implications are enormous.  At stake, potentially, are the freedoms we take for granted under the Bill of Rights; the powers of the president, Congress and the courts; and the policies the government can or cannot pursue.  Conventioneers could alter absolutely anything about the way the United States is governed.  Some say they want to terminate all federal taxes and to require super-majorities in the House and the Senate to put any new taxes in their place.  Others want to bar the government from carrying out a number of its functions, for example by constraining its ability to regulate interstate commerce.  Whatever changes a convention approved would be enshrined in the Constitution if three-fourths of the states ratified them.

Yet the processes for impaneling the convention, selecting the delegates, setting the convention’s voting rules, and determining what issues the convention would consider and how much of the Constitution it would seek to rewrite are a mystery.  That means that under a convention, anything goes.  There are no rules, guideposts or procedures in any of these areas. . . .

Click here for the full post.

Ryan’s Call for “Dynamic Scoring” in Tax Reform Would Invite More Mischief

October 3, 2014 at 2:18 pm

“The reality of tax reform . . . is that any politically feasible plan to scale back tax benefits doesn’t generate enough money to significantly cut tax rates without increasing the deficit,” my latest post for U.S. News’ Economic Intelligence notes.  “Rather than grapple with this reality, . . . House Budget Committee Chairman Paul Ryan invoked the last refuge of supply-side tax cutters in recent comments about how to proceed with tax reform.”  Specifically:

Ryan wants to change long-established methods for estimating the revenue effects of proposed tax changes that the Congressional Budget Office and Joint Committee on Taxation use to “score” the budgetary effects of such legislation.  Ryan . . . badly mischaracterizes existing revenue estimation methods while ignoring the fatal flaws in requiring budget crunchers to use so-called dynamic scoring.

Contrary to Ryan’s claim, current revenue estimates reflect many kinds of changes in households’ and business’ behavior resulting from proposed policy changes.  But they don’t reflect possible changes in the overall level of economic activity that might result from proposed legislation — and with good reason:

First, estimates of the macroeconomic effects of tax changes are highly uncertain.  Second, the most credible estimates usually show changes that are quite small.  Finally, and quite importantly, dynamic scoring would impair the credibility of the budget process because the resulting budget estimates will inevitably be controversial and subject to political manipulation.

Adopting dynamic scoring for tax reform, my post concludes, is a gimmick that would only invite more mischief.

Reassessing a View on Federal Accounting

September 26, 2014 at 1:55 pm

So-called “fair-value accounting” is misguided because it would make federal loan and loan guarantee programs look more expensive than they really are, as my colleague Richard Kogan and I have explained.  Jason Delisle and Jason Richwine, writing in the latest issue of National Affairs, correctly note that the logic of our argument is inconsistent with a 2005 CBPP analysis of proposals to invest part of the Social Security trust funds in stocks instead of Treasury bonds.  We concur.  We have re-analyzed our assessment of investing a portion of the Social Security trust fund in equities and now come to a different conclusion than we did in 2005.

The current method of accounting for federal credit programs fully records — on a present-value basis — all the cash flowing into and out of the Treasury.  In contrast, fair-value accounting would add an extra amount to the budgetary cost, based on the fact that loan assets are somewhat less valuable to the private sector than to the government for several reasons: businesses must make a profit; they can’t put themselves at the head of the line when collecting a debt; they borrow at higher interest rates; and private-sector investors are risk-averse — they dislike losses (in this case, higher-than-expected loan defaults) more than they like equal, and equally likely, gains (lower defaults).  None of these factors represents an actual cost that the government incurs when it makes loans.

Including in the budget a cost that the government does not actually pay would overstate spending, deficits, and debt, making the federal budget a less accurate depiction of the nation’s fiscal position.  It would also treat different federal programs inconsistently, because it would not make a similar adjustment for non-credit programs whose costs are also uncertain and variable.  In a recent article, New York University law professor David Kamin thoroughly explains why “including the cost of risk would skew budget estimates.”

Proposals to invest the Social Security trust funds in the stock market raise similar issues.  Stocks produce higher returns than Treasury bonds on average over the years, but they also entail a greater risk of losing money.  That risk is an important consideration in assessing the pros and cons of a proposal, but it’s not an actual cost to the government and therefore doesn’t belong in the budget.  This conclusion differs from the one CBPP reached in 2005, which, upon further consideration, we now believe was mistaken.

(Proposals to replace Social Security with private accounts are very different, since individuals, rather than the government, would bear the risk of holding their retirement savings in stocks.  Individuals are rightly risk-averse.  As a result, any analysis of their well-being — as distinguished from analysis of the impact on government finances — should account for the variability of the stock market.)