Carol Mead, 67, is retired and her husband, 70, must work at a local rock quarry to close the gap from her Social Security. (Brett Carlsen/AP)

Douglas Elmendorf was director of the Congressional Budget Office from 2009 to 2015.

Benefits for older Americans — especially through Social Security and Medicare — account for the largest part of federal spending today and for the lion’s share of the spending growth that will occur in coming decades without changes in policies. That growth is not surprising: With baby boomers moving into retirement, the number of beneficiaries of those programs is surging. Indeed, in the Congressional Budget Office’s current-law projections, all federal spending apart from Social Security, Medicare, defense and interest on the debt will amount to about the same percentage of gross domestic product 25 years from now that it did 25 years ago. At the same time, federal debt is now larger relative to the economy than at almost any point in our history and is on an upward long-term trajectory. Therefore, cuts in Social Security or Medicare benefits, or increases in the taxes used to finance those programs, will almost certainly be needed to put federal debt on a sustainable path.

In deciding what benefit cuts or tax increases would make the most sense, two developments of the past few decades are especially important: First, the incomes of people across most of the income distribution have risen quite slowly, while incomes at the high end have risen rapidly. Most people in the country have seen little increase in their earnings despite gains in total output and income. Second, changes in labor markets are significantly reducing the role of traditional employer-provided retirement benefits. In particular, the number of workers with defined-benefit pension plans has fallen sharply.

If we narrowed the gap between federal revenue and spending through significant across-the-board cuts in Social Security and Medicare, we would significantly reduce total retirement income for many lower- and middle-income people. That approach would be wrong, in my view, because it would impose a large burden on the people who have been experiencing the slowest income growth. It would also be wrong because those benefits have distinctive characteristics that are even more important for people who do not have defined-benefit pensions: In contrast with most personal saving, Social Security benefits protect against longevity risk because they are annuities, they impose no financial-market risk, and they do not require long-term planning or self-control.

To avoid significant across-the-board cuts in Social Security and Medicare benefits, some of the standard arrows in reformers’ quivers should be a last rather than a first resort. For example, an increase in the eligibility age for full retirement benefits in Social Security would reduce monthly benefits for everyone who chose to retire at the same age as they would under current law, and it would reduce the number of years of benefits for everyone who chose to retire later. That change would be especially harmful for lower-income people: A recent study by the National Academies of Sciences estimated that people with incomes in the bottom two quintiles will experience almost no increase in life expectancy over a few decades, and under current law the eligibility age for full benefits is already undergoing a gradual transition from 65 to 67. At the same time, I would not increase Social Security benefits across the board, as some have advocated, because I think scarce federal resources should be used in more targeted ways.

Instead, we should focus on reducing Social Security and Medicare benefits for high-income beneficiaries and raising payroll taxes on workers with high earnings. For example, the Congressional Budget Office has analyzed options to lower Social Security benefits for workers in the top half of the lifetime earnings distribution but leave benefits for those in the bottom half unchanged. There are a variety of ways to increase tax revenue for Social Security by imposing a payroll tax on income above the current-law taxable maximum. Taken together, such targeted measures could eliminate much of the estimated 75-year shortfall in the system. In Medicare, additional tax revenue could be raised by boosting the payroll tax rate for workers with higher earnings. Also, the additional premiums paid by higher-income beneficiaries in Medicare Part B could be extended to cover a much larger share of beneficiaries than the current 6 percent (a figure that will increase in coming years under current law but just by about half a percentage point per year).

To be clear, this approach to improving the financial outlook for Social Security and Medicare — and for the federal government as a whole — does not offer a free lunch. A substantial improvement would be achieved only if a substantial number of people with higher incomes received lower benefits or paid higher taxes, which would reduce people’s incentives to work and save. Moreover, making Social Security and Medicare more progressive would weaken the connection between an individual’s taxes and benefits, which could undermine the important earned-benefit character of the programs if taken too far.

But the key question, as is often the case in public policy, is whether there is a better alternative. It would be extremely difficult to narrow the projected gap between federal spending and revenue without changing either benefits or taxes for the largest, fastest-growing programs. Changing Social Security and Medicare through across-the-board cuts would impose much of the burden on people who have fared the worst in economic terms in recent years. Instead, we should protect people of modest means by imposing most of the burden of changes on those who are more affluent.