Social Security was designed to be self-financing. Taxes collected from today’s workers pay for benefits to today’s retirees, disabled workers, and their dependents. So in theory it shouldn’t affect the rest of the federal budget.

But the amount collected never exactly equals the amount paid out in benefits, and where the extra money goes to and comes from affects the fiscal cliff. When Social Security collects more than it pays, the excess is invested in interest-bearing U.S. Treasury securities. The government uses the cash from the securities for other things, such as paying down the federal debt, reducing other taxes, or providing other government services. From the mid-1980s to 2008, taxes dwarfed benefits, creating a trust fund now worth more than $2.7 trillion. The trust fund earns more than $100 billion a year in interest, which is reinvested in securities.

When Social Security taxes fall short of benefits, some of the securities in the trust fund are redeemed to make up the difference. That’s what’s been happening since 2009, as high unemployment slashed tax revenues and baby boomers began retiring. In the fiscal year that ended September 2012, Social Security paid out $48 billion more than it collected. The $100 billion in trust fund interest more than covered the shortfall, so for now the trust fund continues to grow. Unfortunately, the shortfall will also continue to grow, reaching $115 billion a year in 2020 and $295 billion in 2025 according to Social Security trustee projections, until the trust fund is exhausted in the early 2030s.

Money is diverted from the federal government’s general revenues each time Social Security collects on an IOU. Closing the gap between what the program is taking in and paying out would reduce the federal government’s deficit.

Many people don’t think so, especially because low and moderate earners pay a disproportionate share. Unlike the income tax, the flat Social Security payroll tax levied on employers and their employees doesn’t exempt low-income workers, and the tax rate doesn’t increase with earnings. Additionally, only wages and salaries up to a certain limit - $113,700 in 2013 - are taxable. As a result, more than five in six low- and moderate-wage earners pay more Social Security taxes than income taxes.

By law, the $2.7 trillion now in the trust fund can only be used for Social Security payments. But the law does not compel Social Security to deplete the trust fund, and it certainly doesn’t have to empty the trust fund over the next 20 years, which will happen under the existing tax and benefit schedule.

To stimulate consumer spending, the Obama administration and Congress agreed to reduce the employee share of the Social Security payroll tax from 6.2 to 4.2 percent for 2011 and 2012, reducing tax revenues about $200 billion over the two years. (The employer share remained at 6.2 percent.) However, the federal government transferred money from the general fund to the Social Security trust fund to make up the shortfall, leaving Social Security unharmed. Policymakers are debating whether to extend the tax holiday for another year, which would boost workers’ take-home pay but swell the federal deficit.
Some advocates worry that the payroll tax holiday threatens Social Security’s long-term financing. Reducing the payroll tax and relying on congressional allocations to make up the shortfall moves the program away from a self-financing social insurance program with dedicated revenue. It may leave retirement benefits less secure and more vulnerable to the political process. In fact, counting the 2012 general revenue transfers to offset the payroll tax holiday raises Social Security’s drain on federal coffers beyond $150 billion this year, making it a more tempting target for budget cutters.

Aside from the payroll tax holiday, the debate mostly focuses on changing the formula for Social Security’s annual cost-of-living adjustments (COLA). The existing formula ties benefit increases to the change in the consumer price index (CPI-W) so inflation doesn’t erode Social Security checks. However, the CPI-W overstates inflation because it doesn’t fully account for quality improvements in consumer goods or consumers’ ability to shift to less costly substitutes when prices rise. The chained CPI, an alternative measure created by the Bureau of Labor Statistics that better corrects for these problems, generally rises 0.3 percentage points more slowly each year than the CPI-W. That may not sound like much, but it adds up over time. The Social Security trustees project that switching to the chained CPI could reduce benefit payments by about $550 billion over the next 20 years (in constant 2011 dollars).

Those cost savings, of course, come out of benefit checks, and the oldest retirees and long-term disabled workers get hit hardest. Switching to a COLA based on the chained CPI would reduce benefits for an 87-year-old who began collecting at age 62 by about 7 percent, a potentially painful cut considering that half of beneficiaries in their eighties collect less than $12,000 a year.

The chained CPI ignores another problem with the existing COLA: it doesn’t reflect the spending patterns of older Americans. The CPI-W is based on the typical purchases of sampled wage and clerical workers. Older Americans spend more on medical care and housing. The Bureau of Labor Statistics has constructed an experimental price index based on the spending patterns of consumers age 62 and older, but Social Security doesn’t use it. Between 1983 and 2011, that price index rose each year about 0.2 percentage points more rapidly on average than the CPI-W because medical inflation surged, so the chained CPI sometimes understates the price increases that retirees experience.

Other types of benefit cuts wouldn’t save much money in the next few years, because most of those proposals would protect existing retirees and those near retirement. Raising payroll taxes would help bridge the financing gap, but few people are talking about that now.

Social Security reform is inevitable. According to the trustees’ latest projections, the trust fund will run out in 2033; after that, Social Security will be able to fund only about three-quarters of scheduled benefits. A combination of revenue increases and benefits cuts seems likely.

Adjusting the Social Security payroll tax is one option. Subjecting earnings up to about $200,000 a year to the Social Security tax would eliminate more than a third of the long-term financing gap. That ceiling would cover about 90 percent of the nation’s wages, the same share covered in 1983 when Social Security last underwent major changes. Only about 84 percent of wages are now taxed. Taxing all wages would nearly eliminate the entire long-term financing gap.

Other changes being discussed include expanding the tax base to cover other types of compensation (such as the value of the health benefits), raising the retirement age, bringing more state and local government employees into the system, and restructuring the benefit formula to favor low-wage workers even more than under the current system.

Whatever changes Congress chooses, it should decide as soon as possible. Most proposed benefit cuts would protect existing retirees and those near retirement, so the pain can be spread over more generations if reforms are enacted soon, instead of have a single group shoulder the burden alone. And it’s best to give people time to adjust to the new retirement rules, by working longer or saving more, for example, before they take effect.

Realistically, it may not do either. Many of the current proposals to limit deductions would actually make the tax code more complicated, not less. This is because the deductions wouldn’t be fully eliminated; they’d be capped. Such caps could raise significant amounts of revenue, but they don’t move us toward a simpler tax system.

It is true that rolling back deductions could raise a lot of money. But it is exceedingly difficult to see how you could raise the amount of money the president is looking for - $1.6 trillion over 10 years - by doing that alone. This is especially true if, like the president, you want to raise revenues only from folks with high incomes. The more you limit the number of people whose deductions you are willing to cap, the less revenue you can raise.

Another challenge is the pressure to exempt some deductions from possible caps. Strong voices are looking to protect the tax deduction for charitable contributions, for example. If lawmakers decide to keep that deduction in place, they would give up potential revenue - and perhaps encourage other groups to seek protection for other deductions.

In discussing tax policy, you always need to distinguish between short-run and long-run effects. What may make sense for the future may not make sense for the right now.

For example, you can make a convincing argument that many tax deductions ought to be reformed or eliminated down the road. The mortgage interest deduction, in particular, is quite hard to defend. It doesn’t appear to encourage homeownership, but it does encourage going deeper into debt or purchasing a bigger home. Other nations, like Canada and Australia, don’t have a home mortgage interest deduction, and their homeownership rates are as high as ours.

In the short run, though, immediately reducing the mortgage interest deduction could shock the housing market, which still remains fragile. And there are fairness concerns. Somewhere in America a young couple just bought their first home; they sat down and crunched the numbers and were able to make it work because of the deduction they expect to receive. To immediately eliminate the deduction for that couple, or even dramatically reduce it, would seem unfair to many of us. So, we may want to phase out the mortgage interest deduction slowly instead of eliminating it right away.

The deduction for charitable contributions raises a whole other set of issues. There is evidence that the amount some folks - especially high-income folks - give to charity is influenced by the incentive of a federal tax deduction. Most proposals out there, whether they’re capping deductions or limiting the value of contributions, would reduce incentives to give to charity, especially for high-income taxpayers.

No, not necessarily. In fact, workers at all income levels may see a tax increase next year depending on the particulars of any deal.

The Bush-era tax cuts have grabbed the spotlight in recent months, but as my TPC colleagues and I discuss in our report on the fiscal cliff, many other tax increases are scheduled to occur at the end of the year. Most notably, the payroll tax holiday is set to expire. If so, a family with $50,000 in wages would pay an additional $1,000 in payroll taxes.

In addition, we should keep an eye on the tax credit provisions that were part of President Obama’s 2009 stimulus bill - including expansion of the earned income tax credit and the higher education Hope Credit - which target low- and moderate-income families. Those credits will revert to pre-2009 levels on January 1 unless Congress extends them.

The AMT is a particularly interesting aspect of the fiscal cliff debate because it’s an unresolved issue about what people will owe in taxes for 2012. Other expiring provisions will affect 2012 liability, including the deduction for state sales taxes and a number of business incentives, but the AMT will have the biggest impact on the largest number of taxpayers.

The current structure of the individual AMT dates from the 1986 Tax Reform Act and was designed to ensure that high-income earners paid a minimum level of taxes. But, because the threshold for who has to pay the AMT is not indexed for inflation, the number of filers who could be subject to the tax has grown to include middle-income earners. Congress has not permanently corrected this, but since 2001, it has repeatedly put a “patch” on the AMT that temporarily increased the amount of income exempt from AMT in order to prevent too many taxpayers from having to pay it.

The problem we now face is that the AMT has not been patched for 2012. Unless something is done, when people file their 2012 returns, the number paying the AMT will go up from 4 million to 32 million, and current AMT payers will see a large jump in their tax liability. Because withholding tables for tax year 2012 assumed an AMT patch, many of those payers will not have withheld enough tax or made sufficient estimated payments during 2012, so they will have a big and unexpected payment due on April 15 if the patch is not extended. And many of these are middle-income taxpayers, who will be hit by the AMT because they live in high-tax states or have lots of kids.

But the damage of an unpatched AMT goes beyond just the effects on those paying it. The IRS has programmed its systems assuming a patch is in place and has informed Congress that it may not be able to process many 2012 returns until late March if the patch is allowed to expire. About 60 million taxpayers, many of whom will end up with no AMT liability, may be unable to file returns for several months. Unless Congress acts soon, this means that many who normally file early will find their refunds delayed .

Few if any members of Congress want to return to an unpatched AMT - it’s one of the rare things that almost everyone agrees on. But so far no one has publicly proposed a separate patch, so the fate of the AMT hinges on the outcome of the larger political bargaining over taxes and spending.

The term “fiscal cliff” is a catchy metaphor, but we shouldn’t take it to mean that we will go over a cliff precisely. The mandatory spending cuts - sequestration - will be put in place over an extended period. So the major effects of these won’t be felt immediately, although some agencies are now delaying contracting due to uncertainty about next year’s funding. At the same time, people’s federal income tax withholding in their paychecks may not change immediately, since the Treasury can keep withholding at current levels if it feels a deal is imminent. It’s also important to realize that when we see things like the recent CBO report that projects another recession if we went over the cliff - with unemployment rising to 9.1 percent - these figures are based on the assumption that we would go over and stay over.

The biggest immediate effect may be an increase in withholding of 2 percent of earnings for payroll taxes. But, as we discussed earlier, this tax increase may happen even with a deal.

There is also the issue of confidence, which is hard to quantify but still important. This is where we could see something like a shock in January. How will businesses and consumers feel if we have no deal? It isn’t hard to imagine that there would be less confidence, and this could reverberate through the economy. If lawmakers decide that their best strategy is “going over the cliff,” they will have a tough time reassuring Americans that they are making progress and will come together in the new year.

Spending for both programs is certainly growing more than we want it to, but we need to put that growth in the right perspective. Data for the last decade and government forecasts for the coming decade indicate spending growth of about 6 to 7 percent a year for Medicare and somewhat higher for Medicaid. If you look at spending per enrollee, it’s roughly in line with GDP growth, or a little bit lower, which is where budget hawks want it to be. What makes overall Medicare and Medicaid spending exceed economic growth is increased enrollment. In Medicaid’s case, the two recessions during the past decade and the accompanying slow income growth for low- and middle-income Americans is the cause. In Medicare’s case, the large number of retiring baby boomers is driving the enrollment rates up about 3 percent a year. So if you add that 3 percent a year in enrollment growth to another 3 to 4 percent in spending growth per enrollee, you get a budget problem.

The short answer is no. Moving the program to defined contributions, or a Paul Ryan-like voucher program, and simultaneously producing measurable savings would require limiting the value of the voucher, which probably wouldn’t cover the growth in Medicare spending. One frequently referenced voucher plan would set the value of the voucher as the cost of the second least expensive plan in a given geographic area. Those plans are often much less expensive than traditional Medicare, but that does not mean they should drive a budget solution; they can be very small and may not be able to serve a lot of beneficiaries. If traditional Medicare is kept as an option, vouchers based on small, low-cost plans might not cover its costs. So the beneficiary would be stuck having to pay the difference.

A voucher plan would expose beneficiaries to the risk of paying more out of pocket if health costs rise and growth in the value of the voucher doesn’t keep pace. Today’s Medicare program spreads that risk more evenly across the population. If people are interested in opting into private coverage, Medicare Advantage could be adjusted to lower some costs and keep private plans in the system.

The most talked about option is raising the eligibility age, and we will probably have to do this down the road. But if we start raising the eligibility age right now, the people who get dropped by Medicare are going to end up with different insurance arrangements. Some will go onto Medicaid. But with some states considering abstention from Medicaid expansion, a lot of people would become uninsured - even more so if the Affordable Care Act (ACA) is not changed to make those over 65 eligible for the Medicaid expansion. Other people could go onto exchanges, but they haven’t been implemented yet, and they’re not all going to be working well on day one. Raising the eligibility age would also shift costs to employers because people would work longer to stay on their employer’s plans. So it is not as simple a policy as people make it out to be.

A number of other options have been laid out. One is to increase Part B and Part D premiums for middle-income, elderly people. Another is to increase deductibles and make them more rational and uniform across the board. A different approach would be to put an out-of-pocket cap on overall spending, which would reduce the need for people to go out and buy supplementary coverage; you could do that in a way that would provide some savings and still protect low-income beneficiaries. You could also put a premium tax on Medicare supplementary insurance policies that would make Medigap less attractive and free up money to pay the higher premiums. And then there is a whole range of provider payment cuts to prescription drugs, nursing homes, home health hospices, and other providers; in total those could yield some reasonably big savings.

Cutting back Medicaid at the same time the ACA is expanding it creates a real tension. There are not too many options with Medicaid because states have done a good job of controlling cost growth over the last 20 years or so. So the best hope for reining in Medicaid would be to address some of the gimmicks states have used to attract federal dollars. Essentially, states tax health providers, then raise reimbursement rates to bring in federal matching dollars, and then return the money to the providers that paid the tax. The Obama administration has proposed curbing these by requiring states to come up with general fund monies to pay the matching rate. Whether states can do that is unclear, and if they can’t, then the federal government saves money.

There are also some proposals to trim costs associated with dual-eligibles - low-income people over 65 who are generally the sickest of the Medicaid population, are expensive, and often get fragmented and uncoordinated care - through better management. There’s a lot of dollars there, but the savings have never really been demonstrated conclusively. Congressman Ryan had proposed a block grant, but block grants are too stringent and would just shift an awful lot of costs back to the states. Per person caps would not be as disastrous as a block grant, but the design issues are really tough and the data to support such a policy is probably inadequate. We think the best target is provider taxes.

You cannot view Medicare and Medicaid in isolation. Providers have the option of treating Medicaid patients, Medicare patients, or privately insured patients. Medicare and Medicaid pay less for the patients they cover than private insurers do. If providers already get paid more for privately insured patients, you can easily see where Medicare and Medicaid beneficiaries are going to have access problems if reimbursement rates are pushed down.

So, we must be careful when reducing Medicare and Medicaid reimbursement rates to save the federal government money. We have to have federal policies that put pressure on private insurers to keep their costs down too. Those policy options are available, and having a public option in the exchanges or expanding rate setting to private plans in the exchanges could apply that pressure. We could also try to limit the tax exclusion for employer-paid premiums. Doing that would encourage employees and employers to shop for lower-cost plans. You couldn’t eliminate that tax exclusion entirely, but you could limit it in ways that recognize that some high-cost health plans are expensive because they insure a sicker or older population.

We could also look more broadly at controlling private spending. The Independent Payment Advisory Board established by the ACA can’t limit private prices directly, even if that seemed like a good idea. But it can make recommendations to rein in private spending.