Doug Elmendorf: Choices for federal spending and taxes

>> Susan Collins:  I'm Susan Collins, the Joan and Sanford Weill Dean of the Gerald R. Ford School of Public Policy here at the University of Michigan.  And I am delighted to see all of you with us here today.  On behalf of the Ford School community, I'd like to welcome you for another of our series of policy talks at the Ford School.  Today's lecture is co-sponsored by the Department of Economics, and we thank them for their support.  I'm also particularly pleased to have Professor Joel Sillrod, my colleague, here with us as well.  We're delighted and honored to be joined by Dr. Douglas Elmendorf, the Director of the Congressional Budget Office.  As many of you know, the Ford School's mission is to inform and improve public policy.  Doug and the CBO staff are really exemplars of that mission. The charge of the CBO is to provide non-partisan, objective, and timely analysis, to aid the congress in making budgetary and economic decisions.  Well, that is no mean feat, for a variety of reasons, but in particular because of the complexity of our budget, the extreme complexity, and the wide-ranging implications it has for so many dimensions of our economy and its performance, both in the short run and in the long run. But also because the budget, the budget deficit, our debt, and fiscal issues of a variety of stripes are arguably the most partisan issue -- arena -- that we're grappling with at the moment.  So in the face of those partisan pressures and those complexities, Doug and his staff consistently provide the unbiased analysis that really is essential for our understanding the policy options and for informing the debate.  I first met Doug back in the 1990's, quite a long time ago, and as a colleague, former colleague at Harvard University I remember him for his clarity of explanation and his devotion to helping people understand economic issues, especially students.  So in that context -- but of course he's also very distinguished as a researcher. There is a more complete bio in your program and I'm not going go through all of that here, but I did want to note in particular that he's worked at the Council of Economic Advisors, the Treasury Department and the Federal Reserve board, and in those positions among other topics his portfolio has included budget policy, social security, Medicare, healthcare reform, financial markets, and macro economic analysis and forecasting. So that's quite a broad reach.  He's been extremely generous with his time today, meeting with faculty and students at the Ford School and University, and he's graciously agreed to take questions from the audience. So at 1:30 our staff will be going on the isles.  And if you have a question please write it on one of the cards and pass it to the isles by 1:30.  And we will try to include it in our Q and A session at the end.  Professor Rowan Miranda who in addition to teaching a Ford School course on public budgeting is also the associate vice president for finance for the University of Michigan, and he along with one of our graduate students, Brittany Galisdorfer, will select the questions and present them to Doug.  And so with no further ado, it is my great pleasure and honor to welcome Doug Elmendorf to the podium.
^M00:03:38
[ Applause ]
^M00:03:45
>> Doug Elmendorf:  Thank you, Susan, thanks to all of you.  I'm delighted to be here at the University of Michigan today.  I have many friends on the faculty of the Ford School, including Susan, and in the economics department and other places in the university.  It's been great to get a chance to talk with them yesterday and today, and I've been looking forward to the opportunity to talk with all of you about the federal budget.  It has been said, and rightly so, that the federal budget is not just a collection of large numbers.  It's a collection of choices that we as citizens have made about the role that we want to the federal government to play in our society. Unfortunately, the choices that we have made are not mutually consistent.  Going forward, the costs of the federal benefits and services that we're accustomed to receiving from the government will be much higher than the revenues that we're accustomed to sending to the federal government.  And a gap of that size can not persist indefinitely.  We will need to make different choices to bring down spending or increase revenues or do some combination of those two.  And it is those choices for federal spending and taxes that are my topic for today.  I'm going to talk about five questions in particular, the first is how big our projected federal deficit is in debt.  The second, what factors are putting increasing pressure on the federal budget. Third, what are the consequences of rising federal debt. Fourth, what kinds of policy changes could put the budget on a more sustainable path.  And fifth, what criteria might be used to evaluate policy changes. Let's start with the first question, how big are projected deficits and debt.  And the answer is very big.  And here's how big.  If you look at the dark bars at the bottom of the picture, those are the projections that we recently made for the federal budget deficit under current law.  And you can see under current law, deficits actually fall quite sharply from more than 7% of GDP this year to about 1% of GDP by 2015 and beyond. But that's because under current law there will be sharp changes in policies.  Under current law at the end of this year tax cuts enacted since 2001 will expire.  Some extra spending cuts built into law last year will take effect.  So at the beginning of next year there will be a sharp increase in taxes and reduction in federal spending.  That's what pushes deficits down so rapidly. That would be fine if policy-makers really agreed that they wanted these changes and policies to take effect. But in fact, most of the people that I work for don't want large parts of those changes to take effect.  So we prepared for them and for everyone who's interested in this topic an alternative set of projections, not based on current law, but based on what most people might think of as current policies.  Essentially, extending the tax cuts that are scheduled to expire under current law, extending the current rules about spending rather than having these additional cuts.  Under current policies, what we call the alternative fiscal scenario, deficits are much larger, shown by the total height of the bars in this picture.  You can see most of the difference is on the tax side, but also on the spending side.  So under an extrapolation of current policies, deficits would return about 5% of GDP for the coming year, been a trillion dollars a year or ten trillion dollars over the decade.  That's not sustainable.  If we ran deficits that were that large the debt would continue to rise relative to GDP.  This picture shows federal debt from just before the second world war.  You can see they dramatic run-up during the war.  But then a fairly sharp decline in debt as a share of GDP in the '50's and '60's.  You can see where we are now, there's been another great jump in debt relative to GDP in the past few years.  Caused importantly by the financial crisis and the recession and the weak recovery that followed.  But in an underlying way, even without that jump current policies are tending to push up debt as a share of GDP.  We continue those policies, the debt would reach about 90 percent of GDP by the end of the decade.  Not only is that very high by the standards of our history, that's very high by the standards of most advanced countries.  Under current policy, we are moving into territory that is unfamiliar to us, and to most other countries.  That brings us to the second question, which is what are the factors that are putting increasing pressure on the budget.  Essentially, we've had a set of policies that have worked in the past, not quite added up to a no deficit, we've generally run deficits.  But revenues and spending have been fairly close together.  But things are changing.  And what are they?  Well, I think the way to start thinking about the question is to think of the budget in a few big pieces. So this slide shows two bars for each of a broad category of the budget.  And the left-hand bar in each pair is low large that category of spending or revenue has been as a share of GDP on average during the past 40 years.  And the right bar is how large that category will be under current policies as we project them for the end of the decade, for 2020.  So let's look at each of the sets of bars.  The left-hand set of bars is social security and the major healthcare programs, the federal government's healthcare programs are Medicare, which provides health insurance for Americans over the age of 65.  And Medicaid, which provides healthcare for low income Americans of all ages and pays for long-term care, especially for older Americans.  Under current policies that set of programs, which has cost about 7% of GDP on average for the past 40 years will cost about 12% by 2020.  An increase of 5% of GDP.  The second set of bars groups everything else the federal government spends money on, except interest payments.  So it's the defence department, but also all the rest of non-defense government spending.  That whole large collection of programs, and I will come back and talk about the details of this category in a few minutes, but that whole collection of programs is on track to be not the 11-and-a-half percent of GDP that's been on average over the last 40 years, but about 8% of GDP.  So under current policies we're seeing a very sharp shift in the focus of the government spending from a broad collection of programs to this handful of very large and very important programs.  But observe here that the savings in the second set of programs under current policies is not 5% of GDP, it's about 3-and-a-half percent of GDP. So there's an overall increase in non [Inaudible] spending by the government.  In addition to that, as you see the third set of bars, there will be more debt interest payments because the debt will be larger.  So total spending shown in the fourth set of bars will be about 2% of GDP more than it's been historically.  And on the revenue side, revenues have averaged about 18% of GDP.  I'll show you later that there's a lot of variation around that average, but no real trend.
^M00:11:15
It was about that share of GDP when I was born 50 years ago as it was a few years ago.  Under current policies if we extend the expiring tax provisions, and I'll be more specific about what I mean by that later.  Revenues would remain about 18% of GDP. So higher spending on a handful of programs, somewhat lower spending on a wide collection of other programs and revenues about the same leads to deficits that are about 5% of GDP.  That's the level that I said is unsustainable because it just pushes debt up, ever-upward as a share of GDP.  And I'm focussing in this talk on 2020 for concreteness, but you should understand, and I'll talk about the factors underlying this -- these patterns, those factors don't reverse.  If anything they will continue in future years.  So this is a longer-term problem, not just a problem through 2020. Now we cannot go back to the past combination of tax and spending policies.  And this is crucial to understand. And the reason is that most of the change, not all, most of the change in the cost of social security and Medicare and Medicare come from underlying changes in our society and in the economy.  And in particular, those existing federal programs are becoming much more expensive because of the aging of the population and rising costs for healthcare.  And to give you a sense of that, with the retirement of the baby boom generation now occurring there will be in this country in ten years about one third more Americans over the age of 65 than there are today.  On the second point, the rising cost for healthcare, you should understand that the cost of healthcare per beneficiary in this country has risen more than one-and-a-half percentage points more per year than GDP per person.  There have been variations in that as well, but having a substantial gap between the rate of increase in healthcare cost and the rate of increase in GDP is a long-standing phenomenon in our country.  In addition to those underlying factors, of course, we have expanded federal healthcare programs during the past several years.  In particular, adding a prescription drug benefit to Medicare in legislation in 2003, and greatly increasing subsidies in healthcare for lower and middle income Americans in legislation in 2010.  But even without those important changes in law there would be substantial growth in the cost of these programs. And as I I've said, outlays for those programs will total about 12% of GDP in 2020, and increasing beyond that compared with about 7% of GDP on average for the past 40 years.  And that means we cannot go back and repeat the policies of the past.  We will need to change to some new set of policies.  That doesn't tell us, and I will never tell you in this talk, what piece ought to be changed, I want to emphasize that anyway.  Those are choices that we will make as a society.  CBO's job is to provide analysis of the consequences of alternative choices, not to recommend particular choices.  Because those choices ultimately depend on value judgments as well as on the analysis.  Our job in general and what I will do here is talk about the choices that we face, and they are choices in some sense for our elected leaders for whom I work, but more broadly to all of us as citizens of the country.  Now before I talk about possible changes in policies let me first just make sure I cover the consequences of rising federal debt.  I will be brief here, because I think must of this you have been taught, but I want to make sure that we don't just skip over it all together.  Rising federal debt would have significant harmful consequences for the federal budget and the economy.  That's why we can't do it indefinitely.  The first point -- maybe obvious but worth saying -- the more debt that we have the higher the interest payments will be on that debt.  And all else equal then, we will need to cut some other spending or raise tax revenue.  The second problem with rising debt is that the more the government borrows the less money is available from our own saving to pay for capital investment in this country or to purchase assets overseas.  So we, by doing less national saving, we will have less output and income than we will have otherwise. A third consequence of rising federal debt is that it puts limits on policy maker's ability to use tax and spending policies to respond to unexpected challenges. Such as economic downturns or financial crisis.  About four or five years ago debt held by the public in this country was 35 to 40% of GDP.  It's now about 70% of GDP.  If we were to encounter conditions like those we encountered in 2007 and '08 and '09 again starting from a much higher share of GDP -- debt relative to our GDP it would be much harder to develop effective policy responses.  And the fourth consequence of rising federal debt is an increase in the likelihood of a fiscal crisis.  By which I mean a point at which people lose confidence in the government's ability to manage its financial affairs and the government thereby loses its ability to borrow at affordable interest rates.  And we can see such fiscal crisis occurring in some other countries today, we can see a very serious harm that does to those countries.  Okay, so that takes -- that takes us to the fourth question, and I will spend longer here.  This is what kinds of policy changes could lead to a more sustainable budget path.  And I'll spend for him here because this is really central to understanding the nature of the challenge that we face.  To start with, it's not clear just how much we need to reduce deficits.  What is clear, as I said, debt cannot continue to increase relative to GDP indefinitely.  We might chose to have debt flatten out as a share of GDP or we might choose to push debt back down as a share of GDP.  And economics can help you understand the pros and cons of those different choices, but it can't really resolve which is the right one.  I offered three illustrative goals here.  One is to reduce the deficit in 2020, in that year alone, by a trillion dollars.  If we did that, that would roughly balance the budget in 2020.  And with no deficit in that year, the debt would be declining fairly rapidly relative to the GDP, pushing us back more quickly toward the historical levels of debt.  An alternative at the bottom of the screen is to reduce the deficit by $500 billion in 2020.  That would roughly keep debt the same share of GDP in 2020 that it will be next year.  So on average over the decade, holding debt about flat as a share of GDP.  The intermediate option here, the $750 billion of deficit reduction, would be between those two, of course, in terms of its effect on the debt.  It would basically push the path of debt back down to what would happen under the current law baseline.  The picture I showed at the very beginning, which shows that coming down gradually as a share of the GDP.  I hope that was obvious, all three of them, very large numbers.  And we're going to see how large when we talk about the pieces of the budget and what it might do in particular pieces to achieve savings of those magnitudes.  And here's how I would summarize it.  Again, I would start with three broad categories of changes before we get into more details.  Putting the federal budget on a path that is more likely to be sustainable [Inaudible] occur otherwise under current policies, would require us to change current policies in at least one of the following ways.  One is to make major reductions relative to current policies in the benefits that people receive when they get older.  As I was discuss in a moment, most of the spending in that first set of bars I showed earlier for social security and the major healthcare programs go to older Americans.  Second possibility is to substantially reduce the role of the rest of the federal government relative to the size of the economy. Beyond the reductions that are already projected to occur.  And the third is to raise revenue significantly above their historical average as a percentage of GDP. And of course, we could do just one or two or three of those.  You'll notice in each bullet point there's an adjective or adverb indicating magnitude.  Major, substantially, significantly.  That's not an accident. That's the magnitude required to meet any of those targets for deficit reduction that I talked about in a previous slide.  So let's look a little more closely at these different programs.  This slide covers the programs included in that first pair of bars from before.  Social security is projected to rise as a share of GDP.  That's basically because there will be an increase in the share of Americans receiving social security benefits.  About 80% of social security benefits go to retired workers or their dependents or survivors.  The remaining 20% go to disabled workers and their dependents.  Spending for Medicare, as I said, provides subsidized health insurance for Americans who are 65 or older is shown in the second line.  We project that to rise more rapidly than social security in this decade and beyond because it's being pushed up both by the aging of the population, that effects social security, but also by rising cost of healthcare per beneficiary.  The bottom line is more complicated.  It shows spending for medicate and something called chip, which is the children's health insurance program, and something we call exchange subsidies and related spending.  Medicaid and chip current provide subsidized healthcare for certain low income people.  Medicaid also provided long-term care.  And the importance of that should not be underestimated.  About 40% of nursing home bills in this country are paid by Medicaid.  And that's a combination of federal spending and state spending on that program.  It's a joint federal and state program. In addition, under the Affordable Care Act, enacted in 2010, there will be a substantial expansion in El visibility for Medicaid for low income people, and other people with not quite so low income will become eligible for subsidies to be provided through health insurance marketplaces known as exchanges.  That bottom line jumps up over the next decade because it's pushed up by the aging of the population and rising costs for healthcare, and the legislative expansions in eligibility for those subsidies.  I want to emphasize that even with the large expansion of health insurance subsidies for lower income Americans, a great deal of the money in the federal health programs will continue to go to older people with smaller shares to disabled people and non-elderly able-bodied people.  Specifically, if you take the set of programs, health programs, Medicare and the ones in the bottom line, by 2020 about half of that spending will go to Americans over the age of 65.  About a quarter to disabled Americans.  And about a quarter to able-bodied, non-elderly Americans.  That's why I say that if one wants to address the budget deficit through substantial changes in spending in these programs that is likely to mean substantial changes in the benefits provided to older Americans.
^M00:23:22 There are many potential policy changes that could effect spending, cut spending or increase spending, of course, on these programs.  We at CBO prepare menus of options for the congress in response to their questions.  But let me give you a sense of how much you would need to do to add up to a pile of money.  Consider the following five changes.  Gradually increasing the eligibility age for Medicare.  Lowering the annual cost of living adjustments for social security by a little bit. Linking initial social security benefits to the wages in the economy rather than to prices.  Cutting federal Medicaid payments for long-term care by a third.  And raising the basic Medicare premium for doctors, part B doctors insurance in Medicare from 25% to 35%.  All five of those things together would reduce deficits by about $200 billion in 2020.  Each of those is a very significant change in policy with advantages and disadvantages.  As it happens, the five I listed would mostly tend to grow rapidly as a share of GDP.  So they'd do even more for deficit reduction later.  But if you're thinking about those three possible targets I talked about earlier you need to recognize that $200 billion, as large as it is, is less than half of the smallest of all of those targets.  Okay, so that's this handful of large and important programs.  Now let's look at the rest of federal spending.  I dividing the rest here into three categories again, and the sum of these three categories were those bars I showed, the second set of bars in the earlier slide.  The bottom line here called other mandatory spending, that's benefit programs apart from social security and the healthcare programs. So it is unemployment insurance, food stamps, some veteran's benefits, government retiree benefits, and a whole collection of other programs.  That spending is projected.  You can see in the picture it spiked up ward in the recession and because of the recovery act that was passed in, has come back down sharply, and is scheduled under current policies would decline over the course of the decade as the economy improves.  You can see by the end of that line, on the right-hand side, in 2020, well, 2020 is not quite at the end.  But near the end in 2020 you can see that spending would be a smaller share of GDP than it's been at almost any point in the past 40 years.  The other two lines, [Inaudible] defense and non defense, refer to what inside the beltway is called discretionary spending.  That just means that congress votes on the allocation of that spending each year.  The benefit programs work differently, congress puts some rules in place and roughly speaking, whoever turns up, who is eligible, can receive benefits.  I mean, there are a lot of complications around that, but that's roughly how to think about what's called mandatory spending here.  Discretionary spending is spending that the congress picks each year.  Except at the moment we're living under a set of caps that congress put into law last summer.  Under the caps put in place last summer spending on defense and all of these non defense purposes would also decline as a share of GDP over the coming decade.  And would decline to a point at or below the lowest share of GDP in the past 40 years.  And you can see then from this set of end points of these lines on the right why we found the result I showed in that bar graph, which is that all these programs of the federal government are already scheduled to consume a smaller share of GDP.  Much smaller share of GDP than they have on average over the past 40 years. We should [Inaudible] a little deeper into discretionary spending because it covers so many different areas. This slide, just to give you a sense of what's going on a little more granular way, shows defense and non defense funding for this year that's about to end, fiscal year 2012.  Defense funding was $670 billion. People who are concerned about defense spending sometimes focus on the procurement of large weapons and you can see that is about a fifth of spending.  Much more of defense spending is on operation and maintenance and military personnel.  That basically covers the payments to people in uniform, the payments to civilian employees, basic supplies.  So that two thirds of spending in those two categories really cover the day-to-day operations of the armed forces.  The non-defense discretionary spending is most everything else the government does.  And there are more granular versions of this chart that we look at, but to give you a sense of it, one big bucket is education and training. Much of that money goes to state governments or in grants to students.  There's transportation spending like highway construction, income security, some programs for low-income people, veterans benefits, this is mostly veterans healthcare provided through the Veteran's Health Administration.  Health research, about half of that health block is research grants, international affairs and other things.  Under current policies, under the lines I showed you before, previous slide, where spending in these categories falls as a share of GDP, it's actually falling in real terms, in inflation-adjusted terms.  So already under current policies there will be less in the way of services provided in the future than is being provided now.  And if we were to try to address budget deficits through further cuts in this area we would have to agree to further reductions in the services to be provided.  And of course that's possible, and again we give options to congress to do that.  Because you recognize the starting point, in a sense, is already a reduced role for those programs as a share of GDP.  Well, that brings us to revenues, the other big piece of the budget, the whole other side of the budget.  This picture shows federal revenues as a share of GDP over time, as I mentioned before.  It's been about flat on average, around 18% of GDP for the past 40 years with some significant variation around that line due both to changes in the economy and changes in law.  Under the alternative fiscal scenario, this is the extension of current policies, revenues stay close to 18% of GDP.  And to be a little more precise about what I mean there, we're talking about an extension of the tax cuts announced in 2001 and 2003 and 2009, which were extended two years ago for two years, and are about to expire under current law.  The alternative scenario extends them.  There's also a part of the individual income tax called the alternative minimum tax that today cover -- that most of the past few years has covered a few million households. But it's set up in a way without legislative changes it will under current law cover many more people. Alternative scenario assumes that congress continues to extend these temporary fixes they call them that have limited the reach of the alternative tax.  The alternative scenario also extends a whole grab bag of other expiring tax provisions.  Under current law, with expirations all take effect, you can see that revenues jump up as a share of GDP over the next few years, aided also by an improving economy, we think, and revenues would be much higher.  Now there are different ways one might choose to raise revenues to reduce the deficit, if one wanted to.  One direction that's attracted a great deal of interest in the past few years in the policy world, but has long been a focus of analysis, is something called tax expenditures.  Tax expenditures are called tax expenditures because they resemble government programs, government spending programs, in the sense of providing assistance to specific groups of people or for specific activities.  So well-known tax expenditure is the tax deduction for mortgage interest payments, the tax deduction for charitable contributions, which is very important for this place and many others.  The tax deduction for state and local income taxes paid and many others.  Tax expenditures are large.  What the bar here shows is the amount of revenue foregone, loosely speaking.  There are some technical issues here that we can talk about.  But tax expenditures viewed as a sort of spending program would actually be larger than any spending program, including social security, which is the largest single program.  And [Inaudible] pretty large share of total tax revenues, shown in the left-hand bars.  But of course -- well, I should say that many analysts talk about reducing tax expenditures as a way of making the tax code more efficient and maybe more fair.  But of course people who use those tax expenditures like to use them, and taking them away will have costs for them.  And as with spending, modest changes don't get you very far toward the total amount of deficit reduction required to put the budget on a sustainable path.  So for example, if we gradually eliminated both the mortgage interest deduction and the deduction for state and local taxes we could raise about $220 billion in revenue in 2020.  And as with that set of policies a described on the spending side, worth about that money, there are advantages and disadvantages of those changes.  $200 billion is no doubt a whole lot of money, but it is, as I said, not even halfway to the smallest of those targets for deficit reduction that I talked about.  Another approach to raising revenues is to allow the tax cuts that are scheduled to expire to expire.  And as I showed, I'll just hop back up, if you can see if one let all of those tax revisions expire revenue would go up by quite a bit.  If one simply extended the tax cuts on individuals that are due to expire one could raise about $500 billion in 2020.  That is a very large under, that's about the size of that small target for deficit reduction.  An alternative that's discussed a lot is to allow those tax cuts to expire for only income above some threshold, most often discussed threshold is $250,000 of income for a married couple.  That alternative of letting the -- extending the tax cuts except for income above some threshold would raise about $100 billion in revenue in 2020.  Or a fifth or a little less than the effect of allowing all those individual tax cuts to expire.  And that fraction is small simply because most of the revenue that will come from letting those tax cuts to expire would come from people with incomes below that threshold.  As one thinks about different -- whether that sort of change would be desirable from one's perspective, maybe worth noting what tax rates have done over the past 40 years. This picture shows average federal tax rates.  These are not the tax brackets that you might think of.  What we've done in this calculation is take all the taxes people pay, not just the ones that you know you're paying by writing the check to the IRS, but the things like the corporate tax whose burden ultimately falls on some person.  And we've imputed that to particular people.  So the numerator of this calculation is the total taxes paid by people in certain income groups. The denominator is the income that they have, so it's that sort of tax rate.  If you look at the right-hand side of the picture, the last data point here is for 2009, the last year for which we have full data.  You can see a drop in those tax rates in the last few years. That's because of the slowing of the economy, primarily. People's income has fallen.  But as you go back before that last drop, go back to 2006 or '07, you can still see the tax rates have fallen over the past 40 years for all of the income quintiles except the top quintile. That basically reflects a set of changes in federal law, as well as changes in the economy.  I don't have projection of these tax rates if certain changes were made to tax law.  But in general, if you think about the expiration of the tax cuts enacted since 2001 then you might think about tax rates going back to about the level we show in this picture before that break in 2001. The lesson, I think, of this slide and the ones that preceded it on spending was any changes that would be made that would have a significant effect on the deficit would significantly effect a substantial group of Americans.  It is hard to make a set of changes add up to $500 billion, and even harder to $750 or a trillion dollars in a year.  Now as you think about alternative policies of the sort I mentioned, and many, many more that CBO has analyzed and the policy makers talk about, what you should be thinking of is what criteria you might use to evaluate those policy changes.
^M00:36:45
This is my fifth question, and I'll spend a little less time here, but I want to run down a list of criteria. Now the criteria that people use, the weights they attach and how they think about the criteria will vary from person to person depending on the person's value judgments.  I hope as I go through this list you're thinking about criteria you would apply, what's important to you and why, and how that would effect the choices that you would make.  The first criterion is simply how big would the government be.  What do you view as the proper size and scope of the government. Recognizing that over time you cannot pick substantially different sizes of spending and revenues.  With one comes the other, at least over a long period of time. Of course not from year to year, necessarily.  The second criterion that I have here is how will the government's resources be allocated among various priorities.  What do you think the government should spend its money on, who should bear the burden of financing the government's activities.  I'll come back to that distributional question at the end, because it's quite important.  But what -- not just how much should the government do, but what should the government do. And recognizing that the size of the changes needed to bring the budget to a sustainable path can make a pretty big dent in certain functions of the government.  So for example, if we were aiming for that intermediate deficit target of $750 billion and decided to do that entirely through changes in that first set of bars and social security and the major healthcare programs, it would require a one quarter reduction in spending on those programs in 2020.  If instead we focused on that second set of bars, all the government's other spending, it would require a cut in that spending relative to current policies of two-fifths.  If we focused just on revenues it would require an increase in revenues in 2020 of a fifth.  So as one thinks about the priorities one needs to recognize that other parts of the budget that one does not put a high priority on are likely to bear very large cuts or very large increasing, depends on whether they're spending or revenues, and you need to think about how you feel about those categories in addition to the categories that you might most focus on.  And again, there's no right or wrong answer here, it's just a matter of your thinking through the choices.  The third criterion would be how much are deficits reduced in the next ten years ^M00:39:11 and beyond.  We talked about that, there are trade-offs in the amount of reduction. The lower the debt is the better that is for the strength of investment and economic growth over time. On the other hand, the lower the debt is, the more the taxes have been increased or spending about has cut and there are consequences to that.  The fourth criterion would be how would the economic impact be over the short term and the long term.  It's worth spending a moment on this.  The immediate tax increases and spending cuts that will occur at the beginning of next year under current law would be the sharpest tightening of fiscal policy in this country in 40 years.  And given that the economy is already expanding only at a modest pace we think a fiscal tightening of that magnitude would push the economy back into recession.  With declines in output and declines in job, rather than the slow growth of output in jobs that we've seen for the past three years.  On the other hand, if one just delays fiscal tightening as long as possible and allows the debt to mount, then that has economic costs over the medium term and long term, and ultimately becomes unsustainable.  So policy-makers face a difficult trade-off in deciding how quickly to implement policies that reduce the deficit. These sorts of policies I discussed also effect the economy, not only through the amount of government borrowing but through the changes and incentives they give to individuals.  We think about that most often on the tax side.  Changes in tax rates effects people's -- effect people's desire -- work effort and savings.  But also changes in some government benefit programs can effect people's incentives to work and to save.  And the analysis that we do at CBO when we do quantitative estimates of big changes in fiscal policy, we try to track the effects of those changes and incentives as well as the effects and changes of government borrowing. And the fifth criterion is simply who would bear the burden.  Different types of tax increases and spending cuts would effect different groups in different ways. That occurs partly directly through tax checks written or the benefit checks cashed.  But also the changes can effect people indirectly through their impact on the overall economy.  And the different sorts of groups one might care about here could be different income levels, different sorts of changes in policy effect people at the bottom of the income distribution differently from those at the top.  But also there are changes in policy that would effect the well-being of people within income groups.  Changes in tax expenditures, for example, could have very different effects on people in similar income depending on what tax expenditures they have been taking advantage of.  This can also -- another aspect of distribution is distribution across generations. Certain sorts of changes in policy would effect older Americans, others would have more effect on younger Americans.  So a lot of different aspects of this problem that receive, as they should, a good deal of attention.  So let me conclude with this.  To put the federal budget on a path that is more likely to be sustainable than what would occur if current policies were continued law makers will need to adopt a combination of policies that will require people to pay more for their government, to accept less in government benefits and services, or both.  Achieving the amount of deficit reduction necessary to shrink the debt relative to the size of the economy or even to keep it from growing relative to the size of the economy would be a formidable tasks given the large gap of the cost of the benefits and services we've become accustomed to and the revenues we're accustomed to paying.  I'll stop there and happy to take your questions.  Thank you.
^M00:43:06
[ Applause ]
^M00:43:10
>> Brittany Galisdorfer:  Thank you for your remarks. My name is Brittany Galisdorfer, I am an alum of the Ford School, I graduated last year with my master's in public policy and I work now here at the university as a business analyst in business an finance.  The first question is politicians on both sides of the isle tend to point to or dismiss certain CBO reports.  How effective, accurate, and/or objective are CBO projections?
>> That's a great first question.  So first of all, CBO reports are absolutely positively objective.  For 37 years hundreds of people have worked very hard to give the congress analysis that is not influenced by the personal preferences of any analyst at CBO or any directors at CBO.  And we maintain that position very effectively for many, many years before I came to CBO and CBO will maintain it for many, many years after I'm gone as well.  Members of congress understand that, in fact.  There are specific reports that certain members disagree with.  I think every report we issue some members think is completely wrong, and they are not shy about saying that.  And they should not be shy about saying that.  They have every right to raise concerns about the work that we do.  But I think the members also appreciate over time the value of the work that we do, and as much as they may object to a particular study on a particular day, as time goes on and they see the set of work that we do, they are grateful that they had the foresight in 1974 to put us in place.  Now the work -- the value of the work, of course, is higher the more accurate the projections and estimates are.  And projecting is a very hard business.  So our analysis is not always correct, we don't claim that it is.  And in a number of contexts, in fact, we talk about ranges of estimates in order to communicate more effectively to congress the uncertainty that we feel about the work that we do.  In some cases, we can go back and see how accurate our work has been.  When we do economic forecasts, we have a history of our economic forecast to draw upon.  As it turns out, our forecasts are about as accurate as the forecasts made by other economic forecasters.  One might say they're about as inaccurate as the forecasts [Inaudible] the income forecasters.  It is a hard job.  What we try to learn from our past mistakes and to make adjustments, for our projections of spending and taxes, it's harder to know whether we've done the job right or not because at any point in time our projection is a projection under current law.  But congress and the president rarely leave current law alone for a very long time to see what happens.  Also, in most cases, we have -- we would have a projection for a certain law, we'd have a projection of the effects of a change in law, [Inaudible] a new projection.  But if it turns out the new projection is wrong it's hard to know whether that's because the estimate of the policy change was wrong or the estimate of the underlying revenue collections or spending outlays was wrong.  And I'll give you two examples of that in the health area. The cost of the prescription drug benefit added to Medicare a few years ago has turned out to be less than we thought it would be.  And we can tell that, because that program created an entirely new flow of money from the federal government that would have been zero otherwise.  So we had an estimate of what it was, and it's coming to be less than that.  In contrast, in 1997, there were important changes made in Medicare to save money on Medicare.  Costs in Medicare came below what we thought they would.  But there's no real way to know if that's because we underestimated the effects of those changes in 1997 or had overestimated what would happen to Medicare payments without those changes.  In fact, 1997 was following a few years of changes in the US healthcare system, a move toward managed care.  And our own analyst hunch is the thing that we got wrong was not understanding how those changes would have effected Medicare in the absence of the 1997 law changes, rather than misestimating the effects of the law changes themselves.  But there's no way to know that directly. So we do look when we can at the history of our mistakes and the history of our projections, but extracting the lessons is difficult.  And mostly what we rely upon in our work is not just the knowledge and experience of analysts at CBO but the knowledge and perspective of people throughout the analytic community.  Every one of our large reports goes out for external review, and we try to pick people with a range of views at universities or think tanks or in the private sector, to comment on what we do.  Almost all of our estimates are derived partly from conversations with people in various businesses or in analytic positions.  And we draw where we can on that.  So we go to a lot of effort to make our work as accurate as it can be.  But it's also important and we're very clear about this with the congress, that nobody take a particular point estimate with all of its precision that seriously because inevitably there is a confidence region around any analysis that we do.
>> Thank you.  Thank you.  The next question is from Twitter.  Is the Kudlow Report cracked when it says yesterday's CBO report shows ACA is a $6 billion increase on the middle class?
>> So I haven't seen the Kudlow Report, and I don't ever comment on other people's comments on what we've done. I can briefly tell you what we did.  The Affordable Care Act establishes a mandate for many people -- not all -- many people to buy health insurance, and imposed penalties on people who do not obtain health insurance. In 2010, when that act was being completed we did an analysis with our colleagues on the staff of the Joint Committee on Taxation, another congressional support group, of the -- not just of the total amount of payments that would be made through this penalty, but who would make the payments across different pieces of the income distribution.  And we were asked recently to update that analysis to reflect the changes in our baseline projections as well as the effects of the Supreme Court decision, which altered the nature of the expansion of insurance coverage under the Affordable Care Act.  And we released yesterday a short report on who would bear that burden.  I think the total amount in 2016 or 18, picked a particular year of this analysis in the future, and the total amount was about $6 billion, and we showed the breakdown of who would be paying that. And you guys -- I should say, you can see all the work that we do of this sort is available on CBO's web site. We work with congress confidentially when they're developing confidential proposals.  Any analysis that we do of a public proposal or an existing law like the Affordable Care Act is made available to everybody in the congress and everybody with a web browser, essentially, immediately.  So all the reports that we've done are available at CBO.gov.  And you can see read -- when you see something like this or anything else that covers us that you're curious about I would urge you to go read the original source and draw your own interpretation.
>> Yeah.  The next question is, is -- excuse me -- a constitutional balanced budget amendment has been developed -- been declared politically infeasible by many in Washington.  Are there other institutional mechanisms that would compel spending and revenue to align more closely.
>> That's a good question.  We have a report underway that actually has a piece about fiscal rules of which the Constitutional amendment is one example.  And we tried to talk in this part of the report about the experience in the United States and in other countries with fiscal rules.  I think the consensus view, in our view, of that experience is that rules can be a back-stop to help enforce agreements that have been reached on ways to reduce deficits.  But they generally have not been effective at forcing the agreements themselves to be achieved.  So in the late 1980's, there was a law in the US known as Gramm-Rudman-Hulling's, after the three lead sponsors, that set targets for deficits.  An those targets turned out to be much harder to reach than had been anticipated when they were put in place because of economic changes mostly.  And the targets were not reached and the law was thrown out.  So that law was not effective in getting agreement on ways to reduce the deficit.  It was replaced in 1990 by a set of rules that were designed to maintain an agreement that was reached in 1990 by President Bush and the congress.  And those rules put limits on discretionary spending ahead of time, giving a sort of overall amount into which congress had to fit what it was going to appropriate the rest of the year.  And set rules to the rest of the budget, otherwise known as pay go rules.  Or pay as you go, which is to say if you want to raise spending in some way you need to propose some other cut in spending or some increase in taxes that would pay for that.  So you're paying for policy changes as you go. And those rules, the discretionary spending caps and the pay as you go rules are viewed by most analysts as having been fairly effective at discouraging back-sliding from the original agreement.  But they didn't create the agreement, they just tried to keep it in effect for a longer period.  Ultimately, that did not work as the overall budget situation improved very dramatically in the late 1990's, the perceived need for deficit reduction faded.  The immediate need had faded. But the longer term problems were still there.  None the less, the perceived need faded and congress evaded and ultimately threw out some of those rules.  So I think the point about rules is that they work well , they can work well, if is policy makers, ultimately the public, is committed to a certain sort of path.  And this helps to remind people of when they're about to slip off it. But those rules cannot generally put us on the path.
^M00:54:33
>> How does the current deficit effect college students?
>> Well, that's actually a very complicated question.  A few thoughts.  The first is that the effects of a deficit on the economy depend an awful lot on the economic conditions that are in place at the moment. And I'll use a specific example of what people sometimes call the fiscal cliff.  The fact that at the end of this year there will be a big jump in taxes and a cut in spending.  So a sharp narrowing of the budget deficit next year.  But the current economic conditions are such that the constraint on economic activity and on jobs is really the demand for goods and services.  Households are very cautious, businesses are very cautious, state and local governments are spending less.  So all that caution is holding down spending.  And without that spending businesses don't hire, people don't go to work, we're not producing that output.  Under those conditions we think that the tax increases would reduce household spending, the cut in government spending is a direct reduction in the demand for goods and services, so the reduction -- the demand for goods and services layered on top of an economy which already has insufficient demand would lead to such low demand that as I said, the economy would fall back into recession.  So alternatively, if one extended those expiring provisions on the tax side and didn't have the spending cuts, the deficit stayed large next year, that would be better for the economy next year.  But those effects actually switch sign later in the decade, by the end of the decade we'd be better off going off the fiscal cliff because there's so much less government borrowing that as the years mount that reduction in borrowing is worth more for the economy than -- better for the economy -- than what would happen if we just tried to extend all these expiring provisions.  So with this deficit right now means for college students, relative to a smaller deficit, is that it means a stronger economy in the short term.  But a collection of deficits of this size over the years as the economy recovers would ultimately be bad for those future college students, and also bad for current college students who will be at that point in their lives still looking for jobs and trying to find higher income and trying to buy houses and things that will be harder to do with larger deficits.
>> If the cut off age for the healthcare programs you discussed was raised from 65 to 70 or 75, would this be enough to restore healthcare spending to historic levels?
>> I don't have precise estimates at have.  Certainly 75 is a very large increase from 65.  And the way that -- we have some -- so we've done some reports that quantify the effects of changes in eligibility ages for social security and for Medicare on not only spending for those programs but also indirectly on tax revenues.  Because people who worked longer would pay taxes for longer.  We have a nice report from January of this year that does try to quantify that.  I don't have it with me, nor do I have those particular numbers in my head.  I think a few -- but a few considerations to have in mind and this -- for this sort of policy, I mean, one is to remember that the eligible age of 65, which has been entrenched for many years, was set in a time when Americans didn't live nearly as long past age 65 as they do today.  So life expectancy at age 65 in this country is now about half again what it was 40 years ago.  At the same time, of course, social security eligibility age or the normal retirement age is working up.  You can still get benefits at 62, but the normal retirement age that for a long time had been 65 is already on its way to 67.  I think a second consideration that's very important for the healthcare programs is what alternative avenues are available for people to buy health insurance.  So under the Affordable Care Act, there will be a set of these insurance marketplaces, exchanges, established.  If in conjunction with raising the age for Medicare more people were allowed to go -- people between say 65 an some other age were allowed to go into health insurance exchanges and if they were allowed to receive subsidies in those exchanges, that would be a very different situation for them than if the Affordable Care Act is repealed or if it is in place but those people are not made eligible for those exchanges or exchange subsidies. So any sort of change of that sort the consequences depend very much on the rest of government programs and the rest of the health insurance and health care system in which they're operating.
>> It's not a matter of if but when the US collapses from debt and we'll become another Greece, please discuss this briefly.
^M00:59:55
 [ Laughter ]
^M01:00:00
>> I had forgotten about college exam questions.  So that's completely wrong.  It is absolutely a matter of if, not when.  There are a lot of policy changes that we can make that would address the problems I've described. And the last thing you should do is to leave here with a sense that it cannot be done.  It's hard, it will require big changes, and I didn't want to pull any punches about that.  But it absolutely can be done.  The extra burden of those set of programs that I started with, the thing which is pushing up the deficit, I described as 5% of GDP.  Well, that's in some sense only 5% of GDP.  It ought to be possible for society to decide either to change that or to allocate other resources in a way to pay for an extra 5% of GDP. Especially in an economy which in good times will grow 5% over the course of two or three years.  So these changes are all addressable.  And we have fundamentally a much stronger economy than Greece in a whole lot of ways.  So there's nothing inevitable whatsoever about our proceeding down the road to a fiscal crisis.  What is true is that if we don't change current policies, then we will ultimately face a good deal of trouble. But it is not at all inevitable, and you should all leave here -- and everybody who follows this issue should understand is it's not a matter of having no choices.  Greece, in a sense, has no choices.  But we have lots of choices.  We just need to make some.
>> So on the topic of choices, how do you communicate choices to the public?  Currently, they're not in political discourse.
>> That's a hard question.  I spent time before coming to CBO, I was at the Brookings Institution for two years, and one thing I anticipated in there was something called the fiscal wake up tour.  And this was a group of people across a collection of Washington organizations.  The head of The Government Accountability office was the leader of this effort, it had economists from Brookings, from the Heritage Foundation, from the Concord Coalition, which is an anti deficit group.  And the fiscal wake up tour, I played just a small role.  But the group, different members of the cast, if you will, the group visited I think nearly every one of the 50 states over the period of a number of years.  And we'd meet with local editorial boards and with local chambers of commerce, groups of students, to talk about our budget problems.  And the idea was to try to educate a broader collection of people way outside the beltway.  And I liked it a lot, in fact.  I went to some interesting places, met interesting people, had good conversations.  I thought we were doing a good job. But I was reminded in the course of this that it's a very big country.  There are a lot of editorial boards, there are a lot of chambers of commerces.  There are a lot of students, in fact.  And for all the people who were seen, many more were not.  And many of those who we saw were ones who are already interested in budget issues, who read that part of the paper, who had gone to the CBO web site.  So there was a certain amount of preaching to the choir.  And again, I liked doing every bit of it.  But it was a slow road, and -- and so as I said, at CBO now I recognize there are a lot of people in the country who are concerned about these issues and would like to know more.  And others who aren't very concerned except maybe just vaguely in the back of their heads, but would also like to know more, who aren't being reached.  Now in our mission at CBO is really to serve the congress.  We're not -- we're not established and are not funded to engage in general public education.  We do try to make everything we do for the congress available to everyone.  That's why we have this web site.  Before we had a web site we would mail copies of studies to people who wrote to us.  So we try to let it spill over to other people.  But the truth is it's just not a central part of our mission to do that kind of educating.  And we were talking at dinner last night with Dean Collins and others about the general challenge of educating people around the country on topics that are complicated, on which they will not be and never will be and never should be experts.  But none the less need to know enough to make informed decisions at the ballot box.  And I think it is a real -- I think it is truly a great challenge.  And I wish we in fact had the time and resources to do more communicating of what is going on in the budget.  But it's just something I can only do occasionally in response to very nice invitations like this one.
>> Do you see a realistic means to substantially lower the debt without a tax increase.
>> Yes, I talked about a collection of possible changes here.  There are a lot of changes that can be made on the spending side of the budget, a lot of changes that can be made on the tax side of the budget.  I think I've been clear about the magnitude of the gap between the cost of our current programs and the tax revenues that we're used to providing.  And to close that gap enough to stabilize the debt will require very large changes in at least one part of the budget.  And if one focuses on just one part of the budget, then it requires especially large changes in that part.  As I said to you, if you focus just on social security and the healthcare programs to achieve that intermediate target that I discussed to cut in those programs of a quarter.  If you focus just on everything else, it's a cut of two-fifths. If you just raise revenue, it's an increase of a fifth. But those sorts of changes are possible.  We produce options for the commerce that would raise revenues by a fifth, that would cut social security and the healthcare programs by a third, that would cut other programs by another two-fifths.  So as a matter of arithmetic, as a matter of the options that we lay out, there are all spending solutions and all tax solutions, as well as a lot of mixed solutions.  And what choice we make depends on what you and I as a voter tell our elected representatives we want them to do.
>> Why is the intra governmental debt so often not included in the debt to GOP ratio, after all, it is money that is owed by the government.
>> That's a good inside the budget world question.  When CBO talks about the budget we generally talk about something called the unified budget.  A concept invented in the late 1960's out of concern that government budgeting included a whole lot of separate, different funds and accounts that weren't being pulled together, and the commission decided they should be pulled together.  And when CBO came along eight years later it focused and has focused since then on this unified budget.  From a unified budget point of view, money that one part of the government owes to another part of the government doesn't matter.  What matters is the government's relationship with the people of the country.  So I talk about outlays, spending to the county, revenues collected from the country, debt owed to the people of the country and to other countries as well.  But behind all of that, there are a fair number of government funds that have holdings of treasury securities.  And one of the most -- largest of these and one of the most prominent is social security.  Social security has a trust fund, actually has two trust funds. Those funds hold real, honest to God treasury debt that the treasury will certainly pay.  And when social security needs -- isn't collecting enough in its tax -- in the tax revenues to pay for that year's benefits, which will happen at some point in the future under our current policies, then it will come back to the treasury, it will redeem that debt, it will get a dollar it can then use to pay a benefit.  But the rest of the government, the treasury department, when it redeems that debt it has anywhere else to go.  It doesn't have money saved up, in a sense.  Actual benefit payment in that year can only be made out of what will be revenues in that year or some other cut in spending in that year, or borrowing in that year.  So the fact that the social security trust fund is building up some amount of money doesn't actually change the situation for the government as a whole in 2020 or 2030 or 2040.  What matters for the government as a whole later on is the benefits the government will pay, then, and the taxes the government will collect.  So we and most analysts focus on this debt to the public, extracting from all the transactions within the government.  Because we think that's the most useful way to capture the government's overall burden. And it's the debt held by the public which is crowding out other uses of investable money, which is crowded out investment and reducing capital stock and reducing incomes over time. ^M01:09:35 So we do report on some more -- larger aggregates of government debt, but we definitely give them much less weight because we think they're much less relevant to the decisions that congress faces.
>> This is the last question.  If all governments are operating with record deficits, won't we in the long term pay for it with inflation.
>> So in the past, in some countries that have run truly large deficits, they have applied pressure to their central banks to finance those deficits.  More or less, literally by printing money.  And in those countries when that process gets out of hand, sometimes people have encountered the phenomenon called hyper inflation, which is very, very high rates of inflation.  And you end up with currency with many, many zeros after it.  I don't think that is the future path for the US.  We have independent central bank for whom a critical part of the mandate is to keep inflation low.  And I think the pressures that the government will face over time will not come out in the form of that incredibly lose monetary policy that has happened in some other countries.  Instead, it will turn up as higher interest rates, as the government competes for borrowing -- trying to borrow so much it's competing more with private borrowers.  So I think the economic consequence that will be most visible will be an increase in interest rates.  It's also true that over time this high level borrowing, as I say, diminishes national saving, diminishes the investment in equipment and structures to make the economy stronger later.  That's less visible in a way, because it can happen gradually and because the connection to government borrowing is not always completely clear.  So there are real consequences that won't be visible, and some like interest rate increases that I think will be more visible if policies aren't changed.  But I don't think that super high inflation is actually a significant risk.  Thank you very much, it's been great to be here.
^M01:11:57
[Applause]
^M01:12:01
>> Well, thank you very much, Doug, for your very clear and thoughtful presentation about these fiscal issues, which are not only extremely complicated, as we've heard, but also very important for the country.  I would like to thank both Rowan Miranda and Brittany Galisdorfer for the help with the questions.  I would also like to thank all of you for the wide range of issues that you have brought to our speaker today. Right after we adjourn there is a reception that I hope you'll all join us for just outside of this room.  And with that, to end, please join me in a final thank you to Douglas Elmendorf.  That was wonderful.
^M01:12:45 [Applause]