Jan Hatzius Issues A Correction To Bernanke's Take Of The Goldman Analyst's Report
During his presentation to the Senate yesterday (to be followed promptly by another presentation before Congress shortly), Bernanke discussed the impact of the $61 billion spending cut on GDP. In doing so he referenced a report published by Goldman strategists Jan Hatzius and Alec Phillips. He did so incorrectly. And the first thing Hatzius did this morning is to correct the Chairman: "Some have wondered—e.g. in the Q&A portion of Fed Chairman Bernanke’s monetary policy testimony on Tuesday—how such seemingly small cuts could have such a noticeable impact. But it is important to remember that we are talking about a hit to the quarter-on-quarter annualized growth rate of spending here, not about a hit to the level of GDP. For illustration, it is useful to go through a simplified version of the calculation underlying our estimates for the House-passed spending cut." Hatzius clarifies further: "We estimate that the $25bn cut in our budget projections reduces growth in Q2 by around 1 percentage point (annualized); this effect is already incorporated in our forecast that real GDP will grow 4% (annualized). In addition, we estimated that the $61bn cut passed by the House would reduce growth in Q2 and Q3 by 1½-2 percentage point (annualized) in Q2 and Q3. (In other words, relative to the assumptions currently embedded in our forecast, the House-passed package would imply an additional ½-1 percentage point drag on growth in Q2 and an additional 1½-2 percentage point drag in Q3.) Spending would then be maintained at that lower level thereafter, and the effect on GDP growth would dissipate quickly in Q4 and would be essentially neutral by 2012 Q1." So perhaps the Chairman will keep this in mind as this report is surely reference once again today.
From Goldman Sachs:
Fiscal Restraint: A Question of When, Not If (Hatzius/Phillips)
• Deficit reduction measures appear to be gaining momentum in Congress, with some spending cuts increasingly likely. The House of Representatives has passed a $61 billion (bn) cut for the current fiscal year, and the president has proposed a five-year freeze.
• Resolving the structural budget imbalance is indeed important, as we have argued many times in the past. If left unchecked, chronic primary deficits will lead to an increasing stock of federal debt as a share of GDP, raising long-term interest rates and thereby crowding out private investment. This would reduce growth over the longer term and increase the risk of a fiscal crisis.
• But just as deficit-financed stimulus comes with longer-term fiscal costs, the long-term benefit of fiscal consolidation comes with a temporary downside. We assume in our own budget projections a cut of $25bn to discretionary spending in the current fiscal year (which ends on September 30), and $50bn next year. If implemented in Q2, the $25bn cut would reduce annualized real GDP growth by around 1 percentage point; this is already incorporated in our 4% GDP growth forecast for Q2. We estimate that the House-passed cut of $61bn would have a larger effect of 1½ -2 percentage points (annualized) in Q2 and Q3. In both cases, the growth effects would fade fairly quickly.
• Especially with regard to the House-passed cut of $61bn, some have wondered how such a seemingly small cut could have such a noticeable impact on a $15 trillion economy. But it is important to remember that we are talking about a 1½-2 percentage points hit to the annualized growth rate of GDP over 1-2 quarters, not a hit of “as much as 2%” (or $300bn) to the level of GDP, as some have incorrectly reported. The analysis below contains a detailed illustrative calculation to show how we arrive at our estimates.
The effect of fiscal policy on growth has been an integral part of our forecast for the last two years, in light of the 2009 stimulus package and more recently the extension of the 2001 and 2003 tax cuts, which came along with a new payroll tax cut. (See “The Fiscal Package: A Boost to Growth in 2011,” US Economics Analyst, 10/49, December 10, 2010.) We have estimated that these policies boost growth initially, but eventually result in a drag as spending and/or taxes revert to previous levels. The drag on growth from federal fiscal policy for 2011 had appeared to be mostly neutralized as a result of the fiscal package enacted late last year, but the prospect of spending cuts has raised the issue once again.
Eventually, deficit-financed stimulus will come to an end. Temporary provisions are scheduled to expire at the end of this year or next, and will become a drag on growth if they do. More importantly, the federal government faces a structural primary deficit of 1% to 2% of GDP even after removing the effect of the economic cycle and these temporary provisions, and even with an assumption of slower spending growth and a partial expiration of the 2001/2003 tax cuts.
Left unchecked, chronic primary deficits will lead to an increasing stock of federal debt as a share of GDP, raising long-term interest rates and thereby crowding out private investment. In previous research we have found that a 1 percentage point increase in the debt/GDP ratio increases long-term interest rates by an average of 1-3 basis points (bp), while the same rise in the deficit/GDP ratio leads to a 10-20bp increase in long-term rates. (See “Fiscal Imbalances and Long-Term Interest Rates: How Close a Link?” US Daily, May 5, 2010.)
The eventual effect of sustained fiscal imbalance is slower growth and greater risk of a fiscal crisis. Our estimates suggest that a 10-point increase in the debt/GDP ratio lowers growth four years later by 0.2 percentage point, and increases the probability of a debt crisis by 2.5% in the aftermath of a financial crisis like the recent one. (See No Rush for the Exit,” Global Economics Paper, No. 200, June 30, 2010 and “When One Crisis Leads to Another,” US Economics Analyst, 11/04, Jan. 28, 2011.) To avoid this, lawmakers must begin to identify deficit reduction strategies.
Various fiscal commissions have identified a number of measures to bring the deficit under control, including revenue-generating tax reform, a value-added tax, reforms to Medicare and Social Security, and hard discretionary spending caps. Resolving the long-term budget imbalance is likely to require a combination of these types of measures.
For now, discretionary spending cuts appear to be the most likely source of near-term budgetary savings. The president included a five-year freeze in non-security discretionary spending in his FY2012 budget proposal, and the House of Representatives has approved $61 billion in cuts for the fiscal year ending September 30.
In our own budget deficit projections, we assume that some spending cuts would indeed be enacted. As a placeholder—neither the president’s budget nor the final House package had yet been released—we assumed a $25 billion cut to non-defense discretionary spending in the current fiscal year (off of FY2010 levels) and a $50 billion cut (off of the CBO baseline) for FY2012, followed by a freeze through FY2016. (See “The US Budget Outlook: Better, but Not Good Enough,” US Economics Analyst, 11/05, Feb. 4, 2011.)
A cut of this scale would reduce the cumulative ten-year federal deficit by $475 billion, not including reduced interest expense. Put differently, the chronic primary deficit of 1%-2% of GDP is 0.2% of GDP lower at the end of the decade than it would be otherwise as a result of the cuts. This still leaves plenty of room for improvement, but it is a start.
But just as deficit-financed stimulus comes with longer-term fiscal costs, the long-term benefits of fiscal contraction come with a temporary downside, as we noted in research published last week. (See “Spending Showdowns and Government Shutdowns: Effects on Growth,” US Daily, Feb. 22, 2011.) Federal government spending enters directly into the Commerce Department’s GDP estimates, so unless there is a full offset from other components of GDP a reduction in federal government spending must reduce GDP on impact.
How large is this impact? We estimate that the $25bn cut in our budget projections reduces growth in Q2 by around 1 percentage point (annualized); this effect is already incorporated in our forecast that real GDP will grow 4% (annualized). In addition, we estimated that the $61bn cut passed by the House would reduce growth in Q2 and Q3 by 1½-2 percentage point (annualized) in Q2 and Q3. (In other words, relative to the assumptions currently embedded in our forecast, the House-passed package would imply an additional ½-1 percentage point drag on growth in Q2 and an additional 1½-2 percentage point drag in Q3.) Spending would then be maintained at that lower level thereafter, and the effect on GDP growth would dissipate quickly in Q4 and would be essentially neutral by 2012 Q1.
Some have wondered—e.g. in the Q&A portion of Fed Chairman Bernanke’s monetary policy testimony on Tuesday—how such seemingly small cuts could have such a noticeable impact. But it is important to remember that we are talking about a hit to the quarter-on-quarter annualized growth rate of spending here, not about a hit to the level of GDP. For illustration, it is useful to go through a simplified version of the calculation underlying our estimates for the House-passed spending cut.
Suppose that the cut of $61 billion in spending authority in the definition of the federal budget accounts translates into a cut in actual federal government spending in the definition of the national income and product accounts (NIPA) of $15 billion (not annualized) in Q2 and $30 billion (not annualized) in Q3. (In other words, we assume that there is some slippage between cuts in spending authority and cuts in actual NIPA spending. If there is less slippage, the growth impact would be correspondingly larger; if there is more slippage, the growth impact would be correspondingly smaller.) In annualized terms, these numbers are equal to $60bn in Q2 and $120bn in Q3. Suppose also that there are no multiplier effects from the reduction in federal spending. If so, a subtraction of $60bn (annualized) for Q2 and $120bn (annualized) for Q3 from the current level of GDP of $15 trillion (annualized) implies a subtraction of 0.4% from the level of GDP in Q2 and 0.8% from the level of GDP in Q3.
What does this mean for growth rates? The fiscal impact subtracts 0.4% from the level of GDP in Q2; calculated over a one-quarter period (from Q1 to Q2), this is an annualized growth impact of 1.6 percentage points. (Yes, we need to annualize again. We first annualize levels of spending, to compare with levels of GDP which are always presented in annualized terms. We then annualize growth rates, because a 0.4% change in the level of GDP that occurred over one quarter would add up to a 1.6 percentage point change over a full year.) The fiscal impact subtracts 0.8% from the level of GDP in Q3; calculated over a one-quarter period (from Q2 to Q3), this is a further 0.4% below the level in Q2 and is again an annualized growth impact of 1.6 percentage points. Thus, our illustrative calculation implies that the House-passed spending cuts reduce the annualized growth rate of GDP by 1.6 percentage points in both Q2 and Q3.
So what should policy do? While it is clear that the primary deficit will need to be eliminated, when to start the process is open to debate, as is how to generate savings. The optimal time to embark on deficit reduction is determined by balancing two factors. On the one hand, major fiscal consolidation is best left until the economy has entered a self-sustaining recovery. We expect growth of 4.0% in Q2 through Q4 this year, and the recent sharp decline in the unemployment rate, if not reversed, indicates that the time to begin reducing the structural budget deficit is getting closer.
On the other hand, however, the sooner deficit reduction begins, the smaller the amount of restraint that will be required each year to stabilize the debt/GDP ratio on a reasonable timeframe, such as the Toronto G20 agreement to stabilize the ratio by 2016. To accomplish this goal, the structural deficit could be reduced by 1% of GDP per year if started next year, or slightly less if started now. Waiting longer means that eventual cuts will be sharper, or that stated goals will not be achieved on schedule.
One potential solution to the timing problem is to rely on fiscal rules. For instance, imposing hard, enforceable caps on discretionary spending levels for the next few years would likely be viewed by the market as a credible policy, and could increase confidence in the sustainability of the US fiscal position. (See “Squaring the Fiscal Circle: Could a Fiscal Rule Help?” US Economics Analyst, 10/29, July 23, 2010.) Reforms to entitlement programs or tax policy could also help, since once enacted these can phase in over several years and unlike discretionary spending, they do not require annual legislation so are less likely to be changed once enacted.
Ultimately, what goes up must come down. In the case of the federal budget, this means that a deficit-financed boost to growth will eventually lead to a drag. While policymakers can try to smooth the transition by phasing in cuts and incorporating multi-year fiscal commitments, achieving a sustainable fiscal policy will inevitably be a painful but necessary process.