The growing gap in life expectancy raises serious questions about any further increases in the full retirement age, since averages are becoming increasingly misleading when it comes to anything linked to mortality. (Photo: Shutterstock)

(Bloomberg Opinion) –Something significant is happening in Social Security: People are retiring and taking their benefits later. These trends are at least in part the consequence of policy changes made in the early 1980s that were purposefully delayed in their implementation.

Consider this: In 1997, 57 percent of men claiming their retirement benefits under Social Security were 62, the earliest age at which one can do so. By 2017, that share had dropped to 34 percent because more people elected to put off claiming their benefits. As a result, the average age of a new male beneficiary has risen by a full year. (These data exclude disabled workers. There are other ways of doing the calculations, but they all show the same phenomenon.)

Not surprisingly, taking Social Security benefits later is also associated with delayed retirement. According to data from the Current Population Survey tabulated by Courtney Coile of Wellesley College, 38 percent of those aged 62 to 64 were working in 1990. By 2017, that share had risen to 53 percent.

Why are people delaying retirement and claiming their Social Security benefits later? One explanation is increased life expectancy and improved health. This probably isn't the full explanation, however, if only because life expectancy changes have been very uneven, as have changes in health status. A second factor is that private pensions have transitioned from defined- benefit plans to defined-contribution ones, and that creates less of an incentive to retire earlier.

Another important reason is changes made way back in 1983, when a Social Security reform following the Greenspan Commission's recommendations gradually raised what is called the full retirement age from 65 to 67. Full implementation of those increases was delayed for almost 20 years, so that the changes wouldn't unnecessarily disrupt retirement plans. The ongoing retirement age changes have attracted very little recent public attention, even though the first tranche of the rise, to age 66, is already over, and we are almost halfway through the transition to 67, which will be complete by 2022.

The rise in the full retirement age has two types of effects on when people decide to claim their benefits. The first is that it reduces the monthly benefit received at any given age. For example, when the full retirement age is 67, those deciding to take their benefits at 62 can do so only by accepting a 30 percent discount per month on the “full” benefit. When the full retirement age was 65, the discount was 20 percent. (The goal of these adjustments is to offset the longer time those claiming their benefits earlier will receive them, on average. Most of the evidence, however, suggests that many people claim their benefits too soon and would be better off, in expected value, if they waited.)

The second, and more important, effect is that the full retirement age creates a social norm influencing when people retire and also take their benefit. People may be drawn to the age at which benefits are called “full” even though there's nothing particularly special that happens at that age. Therefore, when the full retirement age is raised, people push back when they take their benefits.

The statistical evidence suggests that the increase in the full retirement age has indeed delayed benefit claiming, most likely because of this anchoring effect.

So reforms enacted in 1983 are having an effect on how people nearing retirement behave today. Three conclusions follow:

First, delayed implementation with gradual phase-ins makes policy reforms more politically viable. The ongoing increase in the full retirement age is currently creating almost no political backlash, because it requires no new vote by Congress and because it is happening gradually. That may provide a pathway forward for other policy changes that raise revenue or reduce other expenditures.

Second, the problem with delayed implementation is that the world can change between enactment and implementation.

Policy makers in 1983, for example, likely did not foresee that life expectancy would stagnate for low- and moderate-income workers in the meanwhile. The growing gap in life expectancy raises serious questions about any further increases in the full retirement age, since averages are becoming increasingly misleading when it comes to anything linked to mortality.

Finally, the key role played by the full retirement age in driving decisions — well beyond what an Economics 101 model would suggest — highlights the need for more policies to take behavioral economics into account.

Over the past decade, governments across the world have begun exploring how to use nudges and norms (Richard Thaler has won the Nobel Prize in economics for work in the area). But while there have been some limited successes, a large potential to better incorporate psychology into public policy remains to be tapped.

Copyright 2018 Bloomberg. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.