CHOOSING THE NATION’S FISCAL FUTURE
Rudolph G. Penner
The Urban Institute, Washington, DC
Co-Chair, Committee on the Fiscal Future of the United States
Division of Behavioral and Social Sciences and Education
National Research Council of The National Academies
The National Academy of Public Administration
U. S. Senate
February 11, 2010
Mr. Chairman, Senator Gregg and other members of the Committee, I would like to thank you for this opportunity to testify on Choosing the Nation’s Fiscal Future, the report of a committee organized by the National Research Council and the National Academy of Public Administration and funded by the John D. and Catherine T. MacArthur Foundation. I co-chaired the committee along with John Palmer of Syracuse University.
Like a number of other reports this one describes the unsustainability of today’s budget policies. The arithmetic of the budget problem is simple. Three programs – Social Security, Medicare, and Medicaid – constitute more than 40 percent of spending other than interest in a normal year and all are growing faster than the economy and tax revenues. At the same time Congress has kept the overall tax burden remarkably constant between 18 and 19 percent of the GDP for most of the past 50 years. The combination of three large rapidly growing programs and a constant tax burden inevitably implies a growing deficit if spending for other government spending programs is held to a constant share of GDP. As the deficit increases, the national debt grows faster and faster, and interest on the debt becomes a budget problem in itself. In the baseline projection used for our study, the debt passes 100 percent of the GDP in the late 2020s and 200 percent shortly after 2040 under the very conservative assumption that interest rates and the rate of economic growth remain constant in the face of rapidly growing deficits. It is, however, highly unlikely that world capital markets will tolerate this sort of fiscal profligacy for a long period of time. The market for our debt would collapse long before 2040.
There are many indications of these long-run problems in the 2011 budget just issued by the administration. Spending for Social Security, Medicare, Medicaid and interest already equal almost 70 percent of revenues in 2011. Although the absolute level of the deficit declines from 2010 to 2014, it rises thereafter and begins to grow as a share of GDP after 2018. The debt in the hands of the public rises from 53.0 to 77.2 percent of GDP between 2009 and 2020. The interest bill more than quadruples over the same time period.
Our committee believes that Congress should set a target for the debt-GDP ratio and not exceed it. Given an explicit target the American people could judge how well the Congress and administration are doing in their pursuit of fiscal responsibility. We believe further that a prudent target would hold the debt to 60 percent of GDP. That ratio should be achieved by 2022 and we should begin implementing the necessary policies by 2012. If the nation experiences good fortune while holding the debt to this level, it would be wise to lower the target further.
Admittedly, the choice of 60 percent as a target is a matter of judgment. The committee had to balance the risks of choosing a higher target against the political difficulty involved in getting to something lower. A higher debt-GDP target means running higher deficits. For example, if the GDP grows at 5 percent per year, a 60 percent target implies holding the deficit to 3 percent of GDP whereas an 80 percent target would imply a deficit of 4 percent.
As the target for the deficit and debt-GDP ratio is raised, government draws on a higher proportion of the available supply of domestic and foreign savings and interest rates rise. To the extent that the deficit is financed out of domestic saving there would be less available to finance U. S. private investment and that would mean lower productivity growth than otherwise. Consequently, wage growth and standards of living would also be lower. To the extent that foreign saving is used to finance deficits, Americans would have to devote a growing proportion of their incomes to paying interest and dividends to foreigners and again American living standards would suffer.
A higher debt-GDP ratio also raises the risk of a total meltdown in the bond market. Problems would arise if a recession, wars, or some other emergency pushes the ratio above the target and potential buyers of our bonds begin to doubt our ability to put fiscal policies back on a sustainable path. If people want to see what could happen next, they should look to Ireland and Greece. In the face of a debt crisis, Ireland has been forced to raise taxes quickly and slash spending. Civil service pay has been cut more than 7 percent and social programs have been decimated. Tax increases and spending cuts have amounted to over 5 percent of GDP -- a huge negative stimulus that exceeds the value of our recent positive stimulus program. Greece has not yet decided on very specific actions, but interest rates on its debt have soared relative to German rates and rating agencies have significant lowered their bond rating. Some hope that the EU or IMF will yet bail Greece out, but I suspect that any lender will impose harsh conditions on their fiscal policies. Who would bail us out? The IMF? What would the American people think about having our fiscal policies dictated by a lender?
All this suggests that it might be much better to choose a target for the debt-GDP ratio considerably lower than 60 percent. However, when the committee looked at the sort of policy changes necessary to keep the ratio as low as 60 percent, it concluded that it would be politically implausible to choose a much lower target.
I believe that the most important contribution of our committee was to outline a rich menu of policy options that would get us from here to there. We grouped the options into four packages. At one extreme, the committee asked what spending cuts would be necessary to stabilize the debt-GDP target at 60 percent if the total tax burden was maintained at its historic level between 18 and 19 percent of GDP. That package is called the low spending option. At the other extreme, the committee estimated what tax increases would be necessary to finance currently promised Social Security, Medicare, and Medicaid benefits while other programs grew as determined by current law. There did eventually have to be some slow down in health costs in this package or ultimately the health programs would consume the whole of the GDP. But such a slowdown could be put off for a long time.
Two middle paths were also delineated. They differed primarily in the degree to which benefits were maintained for the elderly population. In the path that was relatively generous to the elderly, spending on infrastructure, research and other types of spending had to be constrained while in the other middle path non-elderly spending could be treated more generously.
The four packages were put forward for illustrative purposes only. The numerous policy options contained in those packages could be put together in an infinite number of combinations.
In the package that avoided any significant increase in the tax burden, the rate of growth of Social Security benefits was held to the level that could be financed with the current payroll tax structure.1 At the same time the actuarial deficit facing the Social Security system was also cured. That required accelerating by 5 years the speed with which the full retirement age reaches 67 and indexing it to longevity thereafter, reducing the indexing of initial benefits for the top 70 percent of earners, and switching to an experimental price index, which has been developed by the Bureau of Labor Statistics and which is expected to grow more slowly in the long run than the current index. In assessing such a package, it is important to differentiate an absolute reduction in the purchasing power of benefits compared to today’s level from a reduction in the rate of growth of benefits. Although the package seems severe, it would more than maintain the purchasing power of today’s level of benefits for all but the most affluent. It would, however, reduce replacement rates considerably below the levels promised by current law.
The rate of growth of health spending in the low spending option had to be held to that caused only by the aging of the population. That is to say, all other causes of excess health care cost growth had to be wiped out.
Two types of cost reducing options are described in our health chapter. One set includes options whose effects on health costs can be estimated by CBO with some degree of confidence. These are options such as increasing Part B and D premiums, increasing the eligibility age for Medicare, and reducing provider reimbursements. The other set involves options whose effects are so difficult to assess that CBO does not provide estimates. These include such initiatives such as using information technology more extensively to track patients and coordinating the treatment of chronic diseases. It would probably involve using every option mentioned in the chapter to some degree to achieve the health spending target of the low spending path. To the degree that the options with an uncertain effect actually worked, the scoreable options could be implemented less painfully.
Another more radical approach to achieving the health cost target would be to put Medicare and Medicaid on fixed budgets. Fixed budgets are used in the universal coverage systems of Canada and the United Kingdom. In Canada every hospital must work on a fixed budget and physicians are limited as to their gross income. Strong political pressures make it almost impossible to keep the growth of the fixed budgets down to the level of GDP growth, but nevertheless the fixed budgets impose some restraint compared to our open-ended budgets for Medicare and Medicaid. The rationing methods that go with the fixed budgets in Canada and the U. K. are anything but transparent.
A different approach to a fixed budget would be to use a voucher system to provide Medicare. The voucher would be used by the elderly and disabled to buy insurance and the value of the voucher would vary inversely with income. It might or might not vary with geographic location and age and it could be combined with changes in insurance regulation to do such things as outlaw the use of pre-existing conditions. Medicaid could be put on a fixed federal budget by shifting to a block grant.
Our analysis clearly shows that there would have to be radical changes in health policy to achieve the target of the low spending path. The pain involved in achieving the target would clearly depend on how much the efficiency of the current system could be improved. Although our health chapter describes a large number of options that purport to do this, we do not claim to have examined every option ever mentioned.
The low spending option also implies severely constraining all other spending as well. The low spending defense path would allow the Pentagon to maintain current personnel policies, but would allow very little investment in new weapons systems. Although it would allow small foreign interventions, nothing as large as the current effort in Iraq and Afghanistan would be possible. All other nondefense spending would have to be lowered considerably below today’s share of the GDP.
In the package that attempts to maintain current law benefits, that is to say, the high spending option, two different financing mechanisms are proposed. In one, the existing income tax system is the primary source of additional revenues and all rates are raised proportionately until the top rate hits 50 percent.2 That happens by 2020. We did not think it prudent to have the top rate go above 50 percent because of the inefficiencies and inequities inherent in the current system.
After the top rate reaches 50 percent, a value added tax of slightly less than one percent is imposed. The value added tax rate must be raised gradually and it reaches 7.7 percent by 2040. In the other approach, the income tax is radically reformed. Almost all tax expenditures would be eliminated; the employer‐provided health exclusion would be capped. There would be two rates. The first, at 10 percent, would start at $22, 475. The top rate of 25 percent would start at $44,950. The rates would rise to 11.1 and 27.7 percent in 2020, but could be lowered thereafter because the elimination of most tax expenditures and in particular the capping of the health insurance exclusion yields growing revenue over time. The initial structure of the simplified tax for 2012 was chosen to emulate the distribution of the tax burden under the current system. It would become gradually less progressive over time, but the Congress could easily fix that if it wished by small changes in the rate structure and standard deduction.
Besides requiring large increases in income tax revenues, the high spending scenario would necessitate a doubling of the Medicare HI tax and considerable increases in the Social Security payroll tax. The payroll tax cap would be gradually raised until it covers 90 percent of earnings; the payroll tax rate would be raised from the current 12.4 percent to 12.7 percent in 2012 and then in steps to 14.7 percent by 2080; and there would have to be a second tier tax that would not earn extra benefits. It would start at 2 percent in 2012 and gradually rise to 5.5 percent in 2060.
By 2040 the tax increases required by the high spending option would raise the overall Federal tax burden by 50 percent compared to the 17.7 percent of 2008 and it would continue to rise after that. I know of no state and local budget projections that go out as far as Federal budget projections, but it is safe to say that if state and local tax burdens are added for comparability, the U. S. total tax burden, which is now considerably below the OECD average, would be higher than today’s OECD average by mid-century.
The intermediate package that has the lower spending path would solve two-thirds of Social Security’s long-run financial problems by cutting benefit growth and one-third by raising payroll tax rates. Solving Social Security’s long-run financial problem also helps lower the unified budget deficit along the way. In this intermediate package, Medicare and Medicaid spending is allowed to grow to 7.2 percent of GDP by 2030 compared to 6.5 percent in the low spending package. Spending, other than that for interest, Social Security, Medicare and Medicaid totals 8.9 percent of GDP compared to 6.8 percent in the low spending path and 10.8 percent in 2008. The increase is devoted to defense and domestic spending on things like research and infrastructure.
The intermediate path with the higher spending devotes a large portion of the spending increase to Social Security, Medicare, and Medicaid. All other non-interest spending is lower than in the first intermediate path.
If revenues for the two intermediate paths are raised using the current income tax structure, the top rate never has to exceed 50 percent. Consequently, there is no need for a value added tax. Of course, a value added tax could be used with any of the four packages to lower the income tax rates necessary for fiscal sustainability.
Although no one believes that changes in the budget process can ensure that the Congress makes the difficult choices necessary to attain fiscal sustainability, our committee felt that there were some reforms that could help the Congress deal with the problem. The main deficiency in the current process is that it is too shortsighted. Most of the effort is concentrated on formulating the budget for the next fiscal year. If the Congress set a long-term goal for the debt-GDP ratio, it would be forced to pay more attention to the long-run impact of policy decisions and it would provide a benchmark for judging whether policies were moving toward sustainability.
Although CBO and OMB make long-run budget projections, those of CBO are produced separately from the Budget and Economic Outlook, which plays an important role in today’s process. OMB’s long-run projections are provided deep in Analytic Perspectives and few readers get that far. We believe that long-run projections should be fully integrated into the main budget documents. Other documents may also be helpful. Australia produces a report every three years examining the effect of budget policies on different generations. Apparently it provokes much public discussion that draws attention to the effect of policies in the long run. In addition, it may be useful to require the president to report every year on the long-run fiscal health of the nation.
The committee discussed how long‐run budget targets might be enforced. Automatic triggers present one option. An automatic sequester of spending was used to enforce Gramm-Rudman-Hollings, but it was not well designed. A trigger cannot impose too much political pain or else Congress will change it. But it should be demanding enough to encourage the adoption of more rational policies.
In conclusion, I would like to thank other members of our committee for the hard work they put into producing our report and also thank the very able staff led by Stevens Redburn who made it all possible. A diverse set of ideologies was represented on the committee, but the group was very congenial and all debates were rational. Few committees are that pleasant.
1. The committee thanks staff members of the Social Security Administration’s Office of the Chief Actuary for providing cost estimates on various proposals contained in the report’s illustrative scenarios and for reviewing text for completeness and accuracy.
2. The Urban Institute–Brookings Tax Policy Center provided the estimates necessary to construct the revenue scenarios underlying the committee’s four policy packages.
An archived webcast of the hearing can be found on the Senate Budget Committee's Web site.