Likely consequences of raising the retirement age in Social Security

Two recent academic articles demonstrate the substantial reaction of the labor market to past increases in the full retirement age in Social Security benefit programs. Although there is some offsetting leakage into disability, unemployment, and welfare programs, and some residual stickiness in work cessation, these studies find that people delay their departure from the labor force This result has strong implications for current discussions in the United States about Social Security reform, as the program faces insolvency within the next 10 years.

As fully and carefully documented by Professors Rabate, Jongen, and Atay (RJA) in an article in a leading economics journal, the Netherlands raised its full retirement age in their basic Social Security program from 65 years in 2012 to 66 years and four months in 2019. Because these increases were step-wise on a birth cohort basis, the researchers employed a regression discontinuity methodology applied to complete administrative data sets of the older Dutch population to study the reaction in the labor market and on public finances.

Even though many Dutch workers retire early owing to generous and flexible employer-provided pensions, the researchers find substantial effects on the employment rate (plus 21 percentage points) as well as substantial effects on participation in other social insurance programs (plus 22 percent, disability insurance in particular) between cohorts facing the old and new full Social Security retirement ages. Despite the increase in costs from other social insurance, the savings on retirement benefits and the increase in tax revenues from greater work activity led to a significant net fiscal gain for the government.

Because of strict labor rules in the Netherlands, with essentially guaranteed employment through the statutory retirement age and mandatory retirement thereafter in most sectors of the economy, the increase in work followed closely to the higher retirement age, with little spillover before or after that age. RJA found, though, that the increase in work was spread across more ages for workers who are self-employed and in sectors with less steep age-wage profiles, conditions which would match large segments of the US job market better.

More generally, RJA emphasized the importance of other government and employer policies that affect employment prior to the statutory retirement age, such as pension early retirement provisions and the generosity and eligibility rules for unemployment and disability insurance programs. Higher preretirement employment rates, such as we have seen in the US in recent decades, increase the effectiveness of and reduce the harm from shifts in the full retirement age. RJA reconcile studies across different countries and times to their findings and show a remarkable degree of consistency in results.

Professors Deshpande, Fadlon, and Gray conducted a broadly similar study of earlier changes in the full retirement age in US Social Security although they had less comprehensive and integrated data sets and missed some factors. They too found increases in labor force participation tied over time to increases in the full retirement age for Social Security benefits, with some stickiness at the old full retirement age (65) which some evidence suggests may be caused by employer conditions. These conditions likely include defined benefit pension plans whose retirement ages did not change for age 65, particularly in the state and local government sectors.

As shown above, an important factor in the extent of increase in work effort from raising the retirement age in Social Security is the ease of getting disability benefits. In the Netherlands, there are no explicit leniencies in eligibility standards for older workers to get disability benefits. By contrast, in the United States, there are several layers of leniencies for workers as young as age 50. But these easier standards reflect an outdated understanding and data about the current labor market and need to be changed. It would be best to do so now, even before the retirement part of Social Security is reformed, in order to maximize the fiscal benefit and minimize any harm that would come to work effort.

There was a time, not long ago, in political discussions around Social Security reform that raising the retirement ages in the program gained a bipartisan consensus, as a natural change in line with increasing life expectancies, improving health, and longer working lives. Now, in these politically polarized times, that political consensus is in doubt. That is a shame, because the evidence supporting such a change is there and strong.

Learn more: Weak Science on SSI | A New and Different Measurement of Changes in Income Inequality | What Happened to PAYGO? | Will the US Population Soon Fall?

The post Likely Consequences of Raising the Retirement Age in Social Security appeared first on American Enterprise Institute - AEI.