The Location of More Than 682,000 Overseas Social Security Beneficiaries

Throughout the year, day after day, my colleagues and I report on immigration matters, dealing with people coming, or wanting to come, to, the United States.

Once a year I write about the reverse: people leaving this country.

While other nations, such as the United Kingdom and New Zealand regularly report on in-migration and out-migration, and thus net immigration, the United States does not do so. The Brits showed 715,000 inward migrants in the year ending March 2020, and 403,000 departures, for a net immigration of 312,000. I do not think our outward migration is as significant, proportionately, as theirs.

I tried to do something about this lack of emigration data during in the Nixon administration; I was then working for a consulting firm and we, along with the distinguished (and late) former Director of the Census Bureau, Richard Scammon, filed a proposal with an arm of NIH that had a competition for migration-related studies. We got the document back in a matter of days, along with a stiff letter saying that the agency would not even consider our proposal or think about the subject. (There was some sensitivity in those days about people leaving for Canada to protest our Vietnam War policies.)

There is, however, one governmental data set that deals with some of the people leaving the States. This population, which grows slowly year after year, consists of those who have opted to receive their Social Security checks overseas. There are rules, of course. You cannot receive your check in Cuba or North Korea; virtually all of the about 432,000 retired workers are U.S. citizens who worked at least 40 quarters (10 years) in the U.S. economy, and must be in their sixties, on retirement, or older. There were also about 242,000 spouses, widows, widowers, or children of the retirees and more than 9,000 disabled workers; 682,888 beneficiaries total as of December 2019.

While 682,888 is a very substantial number, it accounts for only a small fraction of all the people drawing Social Security checks of one kind or another (see Table 5.J11, here). In December 2019, there were 64,064,496 beneficiaries in the Old-Age, Survivors and Disability Insurance (OASDI) system; thus the overseas ones made up 1.07 percent of the total and the other 98.93 percent had U.S. addresses. (The percentage overseas only recently topped 1 percent.)

This is not an economically elite population. Their monthly checks averaged $657 a month, just a hair under half the average benefit for all beneficiaries, $1,383. To an unknown extent, these checks were supplemented by similar checks from other nations' pension systems.

What intrigues me about the global distribution of these benefits is the way that it does not reflect — at all — today's immigration data. The benefit distribution is, among other things, a lagging indicator of immigration to the United States two generations ago. The top of Table 1 shows the nine nations other than Mexico with the most OASDI beneficiaries in 2019; the bottom of Table 1 shows the five largest providers of immigrants in 2018 and, for comparison, Iran. Mexico would be on both lists, but is only listed with the top sending nations here. A ratio of migrants to beneficiaries is calculated for these nations.

Examining the numbers of incoming immigrants in 2018 and the receipt of benefits overseas in December 2019, we find that the ratio for the world is about one migrant to .622 of a beneficiary. That's the average ratio, but the extremes are remarkable. At the top of the list is Japan, where the number of beneficiaries is more than 20 for every incoming migrant. At the other end of the spectrum is Iran, where for each of the four resident beneficiaries there were 2,500 people leaving for the States.

The table, among other things, tells us something about how beneficiaries who know the merits of another nation think about that nation. If one is in the U.S. and has a choice, virtually no one opts to move to Iran.

In an earlier report, we speculated that the kind of distribution shown in Table 1 might indicate a preference among retirees to move to prosperous nations, rather than to the less wealthy ones at the bottom of the table. We also figured that nations with above-average levels of life expectancy, such as Japan, would be high on such lists simply because its residents live longer after retirement than people from other countries.

I noticed another linkage this year. The U.S. and many other countries, but particularly prosperous ones, have totalization agreements; these are international arrangements, just short of treaties, that permit the blending of two nations' pension systems for an individual so that the credits from each nation can be merged, providing a higher level of benefits than would be available otherwise.

For example if a retiree had 30 quarters of credits in Nation A, and another 12 in Nation B, and if each required 40 quarters (or some such measure) the retiree would routinely be ineligible for benefits from either nation. But that person could get a pension if the two sets of credits were combined. These arrangements also produce larger checks than could be expected otherwise. This is probably more than a coincidence; we have totalization agreements with the first nine nations in Table 1, but none with the six at the bottom of the table.

Iran's Negative Distinction. Given the remarkable lack of interest in taking one's Social Security pension from the U.S. to Iran, we decided to look at similar data from other nations in the Caspian Sea basin. This is what we found looking at migrant/beneficiary ratios for Iran and six other nations in that area:

Apparently neither native-born Americans, nor people born around, or near, the Caspian Sea, have any interest in taking their Social Security checks to this region. The highest incidence of such check-receiving patterns is in Turkey, which is near but does not border the Caspian. Even in this case, the incidence of such retirement choices is about one-third of the global average, .209 compared to .622.