From Larry Summers:
Thank you for the opportunity to speak before this committee. You have chosen to address issues relating to austerity at an opportune time as both our economic and our budget situations are in considerable flux and as a broad rethinking of reflexively austere policies is underway worldwide. In my testimony today I want to do three things:
First, I will characterize the economic and fiscal outlook. Second, I will reflect on the economics of austerity, arguing that too little of the policy debate in recent years has focused on the imperative of increasing economic growth which, in the short and medium term, goes back to issues relating to demand. Third, I will comment on some of what I see as policy priorities for the years ahead.
The Economic and Fiscal Outlook
I am increasingly optimistic about our economic recovery. Indeed, I believe our economic prospects now look as sound as at any time in the last 15 years. The late 1990s saw the emergence of a major stock market bubble which was followed by recession in 2001 and slow recovery giving rise to fears of deflation. Soon enough bubbles recurred, this time credit and housing markets, leading me to observe in 2006 and 2007 that again, “The main thing we have to fear is lack of fear itself.” In August of 2007, the financial crisis began with profound distress overtaking the economy in late 2008. Recovery since that time has been real if inadequately paced.
I think it is now reasonable to expect the pace of recovery to accelerate if sound policies are pursued. I base this judgment on a number of considerations.
- It appears that housing has decisively turned with home prices up at double digit rates nationwide over the last year and construction rising sharply. Given that the shortfall in housing construction during the post 2007 bust substantially exceeded the excess inventory created during the bubble period, robust housing demand should be with us for years to come. Strength in housing should also propel recovery through improvements in consumer balance sheets and increased demand for durable goods.
- The United States has the potential to benefit from a substantial renaissance in domestic energy production associated with shale oil, so called “tight oil” more generally and natural gas. It is very plausible that North America will be a net energy exporter by the end of the decade. Increased domestic energy production
will involve investment on a substantial scale approaching $100 billion, and significant job creation, including in hard hit sectors, like construction, and in struggling areas of the country, like Western Pennsylvania. Lower oil and natural gas prices will likely lead to increased consumer spending and to some re-shoring of manufacturing.
- There has been substantial improvement in household, corporate and financial institution balance sheets setting the stage for increased spending. Indeed it has been several decades since household wealth rose as rapidly as it has recently, with both the stock and housing markets providing support.
- While fiscal contraction at an excessive pace has been an important economic headwind since the Recovery Act began phasing out in 2011, most of this blow will have been absorbed by the end of this year, setting the stage for some acceleration in growth unless further policies that immediately reduce demand are enacted. By 2014, it is likely that government will not be a retardant on economic activity for the first time in 4 years.
These favorable aspects of the current situation are real. But optimism needs to be tempered by two unfortunate features of the current situation:
First, the economy has suffered long term damage from the financial crisis and recession of the last several years. Long term, unemployed workers have withdrawn, quite likely permanently, from the workforce. Young workers coming out of school have had much greater difficulty than usual getting on career ladders. Capital investment in new capacity has been held back, as has corporate investment in research and development and the establishment of new brands and product categories. Infrastructure investments on some measures have not kept up with deterioration and obsolescence. It is sobering to contemplate that the CBO estimate of the economy’s potential capacity after full cyclical recovery in 2017 is now fully 7.2 percent or $1.2 trillion below the CBO’s 2007 estimate.
Second, there remain real risks to the recovery. The rest of the world, especially Europe, faces major growth challenges. Increasing inequality acts to hold back spending. There are some signs of froth reappearing in credit markets. Measures of confidence, while improved, remain somewhat depressed and the possibility of geopolitical shocks can hardly be discounted. Everything we know about the aftermath of financial crises from the United States’ 1930s depression to Japan’s experience since 1989, suggests that achieving a return to sustained real growth is very difficult and that premature declarations of victory can be very costly.
Fortunately, relatively good economic news in recent months has been matched by even better budget news. A combination of factors has led to substantial downward revisions in projected budget deficits, to the point where the debt-GDP ratio is now expected to decline through 2020. These factors include a stronger economy, a striking slowdown in the growth rate of health care costs and enhanced revenue collections beyond what might immediately be expected given economic performance. While there are no certainties, experience suggests that favorable revisions in the budget outlook tend to be followed by further favorable revisions and vice-versa. It is therefore reasonable to judge that while the nation continues to face a serious long run fiscal challenge, the budget outlook is today far less grim than it appeared several years ago.
This experience should be a useful caution to all of us involved in policy debates. While it is important to address long run issues, our visibility is limited. For example, the CBO publishes reports that analyze their five-year real GDP growth forecasts versus actual realized growth. Historically, the forecast error is 1.2% per year. To put that number in perspective, it implies that there is about a 1 in 4 chance of that our current estimates of real GDP in 2018 are off by more than a trillion dollars (7.4% of GDP). The error in 10- year projections would be significantly greater.
The Economics of Austerity
Both in the United States and abroad there have, in recent years, been fierce debates about budget policies and ideas around austerity and deficit reduction. These debates which are often framed in universal terms have often shed more heat than light. A prudent government must over time seek to balance spending and revenue collection in a way that assures the sustainability of debts. To do otherwise, leads to instability and needlessly slow growth, and courts default and economic catastrophe. Equally however, responsible fiscal policy requires recognizing that when economies are weak and movements in interest rates are constrained, as has been the case in much of the industrial world in recent years, changes in fiscal policy will have large impacts on economic activity that in turn will affect revenue collections and social support expenditures. In such circumstances, aggressive efforts to rapidly reduce budget deficits may actually backfire as a contracting economy offsets their direct benefits.
It is a truism that deficit finance of government activity is not an alternative to tax finance or to supporting one form of spending by cutting back on another. It is only a means of deferring payment for government spending and, of course, because of interest expenses increasing the burden on taxpayers. Just as a household or business cannot indefinitely increase its debt relative to its income without becoming insolvent, a government cannot either. There is no viable permanent option of spending without raising commensurate revenue. The meaningful choices involve the size of public activity and the timing of government spending and taxation.
It follows that in normal times there is no advantage to deficit policies. Public borrowing does not reduce ultimate tax burdens. It tends to crowd out private borrowing to finance growth and job creating investment and tends to foster international borrowing, which means an excess of imports over exports. Or the expectation of future tax increases may discourage private spending. While government spending, or tax cutting financed by borrowing, creates increased demand in the economy, the Federal Reserve can in normal times achieve this objective by adjusting base interest rates.
It was essentially this logic that drove the measures taken in the late 1980s and in the 1990s, usually on a bipartisan basis, to balance the budget. As a consequence of policy steps taken in 1990, 1993 and 1997, it was possible by the year 2000 for the Treasury to use surplus revenues to retire Federal debt. There is no question in my view that deficit reduction, and the associated reduction in capital costs and increase in investment, was an important contributor to the nation’s very strong economic performance during the 1990s when productivity growth soared and unemployment fell below 4 percent. Essentially, we enjoyed a virtuous circle in which reduced deficits led to lower capital costs and increased confidence, which led to more rapid growth, which further reduced deficits reinforcing the cycle.
As a Treasury official in the 1990s, I was proud to support and help implement these measures. The time will come again when deficit reduction should be the immediate first priority of budget policy.
But, in recent years, circumstances have been anything but normal in the United States and most of the industrial world. High levels of unemployment, low levels of job vacancies and deflationary pressures all indicate that the level of output is not constrained by what the economy is capable of producing, but by the level of demand. Moreover, with base interest rates at or close to zero, the efficacy of monetary policy is circumscribed. In the United States, GDP has been as much as a trillion dollars a year or more than $10,000 per family below its potential.
Under these circumstances, there is every reason to expect that changes in deficit policies will have a direct impact on levels of employment and output in a way that is not normally the case. Borrowing to support spending, either by the government or the private sector, raises demand and therefore increases output and employment above the level they otherwise would have reached. Unlike in normal times, these gains will not be offset by reduced private spending because there is substantial excess capacity in the economy, and cannot easily be achieved via monetary policies because base interest rates have already been reduced to zero. Multiplier effects operate far more strongly during financial crisis economic downturns than in other times.
Two further considerations magnify these effects. As I noted earlier, sustained poor economic performance, in addition to reducing output and employment, adversely affects future economic performance. So, measures that support demand raise future, as well as present, output. Also, support for demand helps to stimulate the economy by offsetting contractionary, deflationary pressures.
In a study published last year in the Brookings Papers on Economic Activity, that I ask be included in the hearing record, Brad Delong and I made estimates suggesting that the effect of expansionary fiscal policies might well be to reduce, rather than increase, future debt burdens because of their positive economic impacts. These estimates remain the subject of substantial debate among economists and I would never want to suggest that policy should be driven by the results of a single study. Yet, I do think it is a fair
conclusion that once account is taken of the direct impact of budget policies on economic performance, their impact on debt burdens is greatly attenuated.
To illustrate: Consider the effect of the sequester in 2013. The sequester will impact the last 10 months of calendar year 2013. The CBO estimates that the sequester will, over this 10 month interval, reduce spending by $64 billion. With no other change, this would result in a reduction of $64 billion in the Federal debt, which is equivalent to reducing the debt/GDP ratio by 0.39 percent.
However, we must also consider the sequester’s effect on GDP growth. The CBO estimates that the sequester will reduce the GDP growth rate in 2013 by 0.6 percentage points. This stifling of growth actually increases the debt/GDP ratio through two effects: First, by reducing the GDP growth rate, the sequester reduces the denominator of the debt/GDP ratio. Second, lower GDP during 2013 means lower tax revenue, which increases the deficit.
We cannot ignore these spillover effects of the sequester onto the economy and onto tax revenue. When we account for these spillover effects, the CBO estimates imply that, the sequester will have a negligible effect on our debt/GDP ratio, at the end of the day.
These observations have strong implications for recent debates over austerity—debates that have reached a crescendo with recent controversies over the work of my Harvard colleagues Carmen Reinhart and Ken Rogoff. I have attached a commentary reflecting my views on their work.
More important than arguments over their data and statistical procedures is the simple observation that the impact of debts and deficits will vary with economic circumstances and the further point that while high levels of debt can retard economic growth, increases in borrowing can enhance economic growth by mitigating downturns. This has the additional impact as I have already noted of raising future potential output.
International comparisons tend to confirm the view that excessively rapid fiscal consolidation has adverse impacts on economic performance.
In Figure 1, we see that countries that pursued harsher austerity policies in recent years also had lower real GDP growth. In Figure 2, we see the difference in unemployment in the US and Eurozone. In 2009, the US and Eurozone had almost the same unemployment rates. In the interim, the Eurozone pursued far harsher austerity policies. Today, the gap in the unemployment rates between the US and Eurozone is 4.6 percentage points.
Naturally, I would be remiss if I did not caution that correlation is not the same as causation. And there are many different ways of processing these data. However, in the face of these data it is difficult to credit claims that more rapid fiscal consolidation is likely to accelerate economic growth.
Policy Going Forward
The foregoing analysis suggests several important principles regarding US fiscal policy in the years ahead.
First, it would not be desirable to undertake further measures to rapidly reduce deficits in the short run. Excessively rapid fiscal consolidation in an economy that is still constrained by lack of demand, and where space for monetary policy action is limited, risks slowing economic expansion at best and halting recovery at worst. Indeed, there is no compelling macroeconomic case for the deficit reduction now being achieved through sequestration, as the adverse impacts of spending cuts on GDP more or less offset their direct impacts in reducing debt. An ultimate judgment on sequestration should therefore depend on a view about the merits of the expenditures being cut back in providing public benefit. I think it is unlikely that aside from the macroeconomic argument, which is dubious, that policymakers would adopt the sequestration cuts simply on grounds of efficient public expenditure though this is not at root an economic judgment.
Second, while uncertainties are great and progress has been made, the United States does face an unsound long run imbalance between forecast expenditures and revenue collections. Spurring growth is the best way to reduce this imbalance. Indeed a 1 percent increase in the growth rate of GDP maintained for 10 years would reduce cumulative deficits by more than $3 trillion. Accelerating growth should be a central aspect of budget debates going forward.
Third, the highest priority in terms of structural reforms to reduce future deficits should be attached to measures that would be desirable even in the absence of prospective deficits. Candidates here include steps to control the growth of health care costs and tax reform. Careful international studies suggest that the excess of US health care costs over foreign costs are more related to a given procedure costing more in the US than to more procedures being performed in the US. This suggests an emphasis on improving approaches to purchasing care rather than on curbing consumer demand for medical assistance. There are a number of features of the tax code that both cost the government revenue and make the economy less efficient. These include corporate tax provisions that support the shifting of economic activity and accounting income to tax havens, subsides that favor particular industries over others, and measures that create an economic bias towards risky financial transactions. Sound loophole-closing tax reform offers the prospect of increased revenues, increased incentives for productive economic activity through lower rates, and increased government revenues.
Fourth, attention should be devoted to measures that reduce future deficits by pulling expenditures forward to the present when they have the additional benefit of increasing demand. It is important to recognize that just as increasing debt burdens future generations, so also does a failure to repair decaying infrastructure, or to invest adequately in funding pensions, or in educating the next generation burdens future generations. Wherever it is possible to reduce future public obligations by spending money today, we should take advantage of this opportunity especially given the very low level of interest rates. In particular, a major effort to upgrade the nation’s infrastructure has the potential to spur economic growth, raise future productive capacity and reduce future deficits. It should be a high priority.
Thank you very much for the opportunity to speak with you today. I look forward to your questions.
See here for speech with graphs. HT: Brad Delong