By Chris Adams

What if a commonly cited statistic was flawed?

For Adam Bee, an economist at the U.S. Census Bureau, tackling that question has been the focus of a major research effort. Bee and Joshua Mitchell were the authors of a recent article that asked “Do Older Americans Have More Income Than We Think?

In a session with National Press Foundation fellows, Bee described the findings of their analysis, which relied on new data sources to gauge whether income that people report on Census surveys is reliable.

Their findings could upend the thinking on the resources available as people move into their retirement years. The reason: Traditional income analyses that rely on the well-respected Current Population Survey Annual Social and Economic Supplement from the Census Bureau are missing income that people actually have at their disposal.

What Bee and Mitchell did was compare Current Population Survey results – based on contacting 75,000 households – with other records, such as those from the IRS and Social Security Administration. They found that many people underreported their income.

Missing was income from defined benefit pensions and retirement account withdrawals.

They don’t say the underreporting is intentional – after all, people likely have the correct, higher number on their tax returns. It could be a matter of forgetting about pension or 401(k) withdrawals, or perhaps not understanding the Census Bureau questions.

Because the traditional Census Bureau numbers are so widely used in public policy debates, the new look at actual income available to seniors is important.

Overall, the analysis found that the median household income from households age 65 and up was 30 percent higher than previously believed. Further, the poverty rate for those older households was 2.1 percentage points lower (6.9 percent, instead of 9 percent).

In his talk, Bee also explained income data sources that journalists and researchers can tap from the Census Bureau or the Federal Reserve’s Survey of Consumer Finances.