Alicia Munnell, head of the Center for Retirement Research at Boston College Alicia Munnell, head of the Center for Retirement Research at Boston College.

“How Much Income Do Retirees Actually Have?” — a provocative, important question and the title of a paper published last month by the Center for Retirement Research at Boston College, whose director is Boston’s management sciences professor Alicia Munnell.

The industry consensus: Americans aren’t saving enough — if anything — for retirement. Some studies find, however, that retirement saving seems to be somewhat on the upswing.

Financial advisors need to keep on top of this issue to help clients plan for what — hopefully — will be their golden years.

“The typical working American has zero, zilch, nothing saved for retirement,” National Institute on Retirement Security executive director Diane Oakley has said.

She was commenting on NIRS research released in September showing that four out of five American workers have less than a year’s income saved in retirement accounts.

The U.S. is tobogganing toward a retirement-funding crisis, according to economist Teresa Ghilarducci, director of the New School’s Retirement Equity Lab.

“I’m shocked that about 20% of people in their 50s making $250,000 a year have nothing but Social Security” for retirement, Ghilarducci told this reporter in a Feb. 8 ThinkAdvisor interview.

Meanwhile, on a positive note, a Bankrate survey out this past August found that 28% of Americans were saving more this year than last. That’s the highest percentage of people who increased their retirement savings since 2011, Bankrate reports.

It was also the fourth year in a row that more Americans put more money away for retirement. Further, about 21% of respondents said they’re comfortable with how their retirement saving was progressing.

In addition, other studies have found that “most households are prepared for retirement,” as Andrew Biggs, a resident scholar at the American Enterprise Institute and former Social Security Administration principal deputy commissioner, wrote in a Forbes article in July.

Biggs cited, for instance, the Urban Institute’s Dynamic Simulation of Income model that projects real median retiree incomes “to rise to $40,880” by 2025 and “to $42,165 by 2035.” In contrast, the median retiree’s income sources and assets in 2015 “support[ed] a total annual income of $37,887.”

How do financial advisors view the crucial and controversial retirement-funding issue?

Harold Evensky, chair of Evensky & Katz/Foldes Financial Wealth Management based in Coral Gables, Florida, says: “I’ve seen research that [shows both sides]. I’m skeptical about studies concluding that in retirement, people will be spending substantially less than [when they were working].

“I think many people are maybe living in a dream world thinking they have enough. I’m extremely concerned,” says Evensky, whose office is in Lubbock, Texas.

Critics of the U.S. Census Bureau’s Current Population Survey (CPS) argue that the poll consistently underestimates retirement income.

To determine whether the findings of four other surveys were in line with the CPS — said to be understating income since the 1980s — the Center for Retirement Research examined income data from the Survey of Consumer Finances, Health and Retirement Study, Panel Study of Income Dynamics and the Survey of Income and Program Participation.

Funded by the Social Security Administration, the CRR’s paper confirms that most people haven’t saved enough for retirement and that “many households are still in danger of not having enough resources in retirement.” The estimates “imply that between 42% and 60% of households [ages 65 and older] are at risk of falling short,” the CRR says.

“We’ve found that the CPS is an outlier and that the other research is accurate at measuring retirement incomes,” says Anqi Chen, CRR assistant director of savings research and an author of the paper.

“Our study shows that we have to be careful when looking at income measurements from [the Census Bureau’s CPS]. Though they’re good at measuring [other demographics],” Chen adds, “it’s been confirmed that they’re not good at measuring income. It’s better to look at other surveys for that.”

She continues. “The study we did was really cool because we [built off a study that was] able to link each person that responded to the CPS to official IRS tax records and official Social Security numbers. We found what researchers have speculated for a few decades now: The respondents to the CPS understate how much [income] they actually have.”

The phrasing of some CPS questions likely accounts for much of that; for example, individuals may not report some withdrawals from their 401(k) plans and IRAs, Chen notes.

For financial advisors, the core of the problem remains: How to effectively encourage clients to save for retirement.

One idea is to substantially cut overhead. That isn’t a new idea — nor apparently one that too many pre-retirees are eager to embrace.

But “once you become an empty nester, you don’t have to remain living in a big cavernous house,” says Keith Collins Reagan, Sr., managing partner of Reagan Holloway wealth management firm in Williamsburg, Virginia. “You can sell the house, buy something much smaller” and invest the balance of the proceeds.

To help avoid an income shortfall once “officially” retired, working at a fun full- or part-time job may be an option. The extra revenue can mean a big reduction in the amount of income drawn down from investment portfolios.

Another approach is to use required minimum distributions (RMDs) from retirement accounts for yearly income planning.

“I recommend to my clients that they start taking their RMD in January and then either monthly or quarterly,” Reagan says. “I advise them: ‘Don’t wait till December to take your RMD. Build your budget around it and use it as a part of your annual income.’”

Evensky frames his advice this way: “People need to do some good planning and be realistic about what they’re going to do in retirement and what it’s going to cost.”

— Related on ThinkAdvisor: