Senator Marco Rubio (R., Fla.) has become the first politician in many years to propose scaling back the future costs of Social Security. President George W. Bush offered his own proposal in 2005, but it never came to a vote in Congress and was widely seen as putting an end to his political momentum following his re-election. Rubio’s proposal will and should be evaluated mostly in terms of its impact on retirement security, national saving, and economic growth. It has modest implications for family structure as well, however, that should not be wholly overlooked.

Public pensions for the elderly began in Germany in the late 1800s, and the German economist Hans-Werner Sinn has noted that a kind of familial decay lurked in the background of their origin. “The idea of having children in order to ensure satisfactory consumption in old age had been common before Bismarck’s reforms,” he writes. Since few people could invest in financial capital, such as stocks and bonds, they invested in human capital: their children. This traditional method of provision for old age—an intergenerational bargain within a family in which parents took care of their children when young and in return received care from them later in life—had several drawbacks. The adult who had never had children, or had children survive to adulthood, or had children with enough income to provide for them, faced a high risk of destitution.

The traditional method of provision for old age—an intergenerational bargain within a family—had several drawbacks.

There was also the risk of the ungrateful child. Sinn writes that Bismarck “mistrusted the benevolence and generosity of the new type of family that had emerged from urbanization and industrialization and saw the pension system as an enforcement device for ensuring a resource transfer to the old generation, one which would not have come about through voluntary private actions.” The various risks involved in the old dispensation means, as Sinn demonstrates, that a “moderately sized” pension system enhances social welfare. If parents are giving back part of their pension in the form of bequests, however, the system has grown too large.

The risk of an overly large system is that it will undermine the incentive to invest in human capital. Public pensions partly socialize the returns to raising children—you don’t have to make the sacrifices entailed in raising them yourself to benefit from their having been raised to a productive adulthood—and therefore result in fewer of them being, er, produced. Several studies suggest that, just as Sinn argues, public pension programs such as Social Security have reduced birth rates by a significant amount in the developed world. (Obviously, nobody claims that they are the only reason birth rates have declined.)

Even more than rising life expectancies, declining fertility rates have made these public pension programs harder to finance. Together they mean that there are fewer working-age adults to support the retirement of the elderly. This demographic difficulty is the backdrop to Sen. Rubio’s proposal.

That proposal has several components. A few of them would encourage people to extend their working lives. Rubio would gradually raise the age at which Social Security benefits begin, lower taxes on the earnings of senior citizens, and ensure that those earnings would not decrease seniors’ benefit levels. He would make it easier for people to save for their own retirements by allowing those without access to company pension plans to participate in the Thrift Savings Plan for federal workers.

Rubio’s changes to Social Security may yield a modest increase in the investment in human capital that is child-rearing.

Perhaps most important, in terms of the program’s solvency, Rubio would change the way future benefits are calculated. His goals in making that change are to extend the life of the program, slow the growth of benefits for high earners, and give low earners more. It is entirely possible to achieve these goals simultaneously because benefits are now calculated in a way that has them grow with time at all income levels. (The retiree of 2045 is supposed to get a check that is significantly higher, even in inflation-adjusted terms, than the retiree of 2015; the bigger check results from the economic growth that should happen in the interim.)

Two of his changes, then, reduce Social Security’s future costs: the increase in the retirement age and the change in benefit levels. These changes should yield an increase in the supply of labor and the stock of capital, as people work, save, and invest more for their old age and rely on the government less. They should also yield a modest increase in the investment in human capital that is child-rearing. They do not eliminate what might be called, borrowing the language of economists, the “free rider” problem of Social Security: The returns to raising children will still be partly socialized. (Not least because Medicare has the same feature.) They will, however, reduce this effect.

For people who would like to see larger families (or see them shrink more slowly), and people who would like government policies to have a smaller effect on family size, this feature of the Rubio plan should serve as an under-appreciated point in its favor.