Today, in honour of Liberation Day, South Korea’s Ministry of Foreign Affairs (MFA) published the 2013 edition of its annual Diplomatic Whitebook. Among other facts and figures, it revealed that 1,991 “reverse migrants” from the U.S. restored their South Korean resident registration and provided proof that they had given up U.S. green cards or citizenship in 2012, slightly down from 2,128 the year before.

This number contrasts sharply with the lists provided by the U.S.’ Internal Revenue Service last year, which contained only 932 names of people of any nationality who gave up U.S. citizenship or green cards. Among those 932, only sixty-three had Korean-language given names — itself the highest total since 1997, but still far below the true number hinted at in other sources. (Ex-green-card holders are only supposed to show up in the IRS’ lists if they held their status “long term”, i.e. for at least eight years; as I discuss later in this post, the IRS seems to think that at least half of ex-green-card-holders meet this criteria.)

Fortunately, thanks to a Freedom of Information Act Request by Shadow Raider, we have another number to which we can compare this: the number of people reported by U.S. Citizenship & Immigration Services to have filed Form I-407 to abandon their U.S. lawful permanent residence status. The MFA’s count of South Korean reverse migrants from the U.S. has generally been 9–13% of USCIS’ count of ex-green-card-holders of all nationalities, and 2012 was no exception. The agreement between these two sources coming from bureaucracies on opposite sides of the globe speaking two different languages — and their common disagreement with the data published by the IRS — is yet another indication that the IRS data is wildly inaccurate.

Table of contents

Introduction to the reverse migration data
Data up to year 2000
Data from 2001 onwards (including USCIS data on green card cancellations)
Analysis: Green-card holders
Analysis: Citizens
Analysis: Falling number of reverse migrants from 1994 to 2008
Conclusion

Introduction to the reverse migration data

Every year since 1990, South Korea’s Ministry of Foreign Affairs (previously known as the Ministry of Foreign Affairs and Trade before President Park Geun-hye’s reshuffle of ministerial duties) has included in its annual Diplomatic Whitebook statistics on the number of “reverse migrants” — South Korean citizens who emigrated in previous years but then made a declaration of permanent return under Overseas Migration Act § 12 in the year under examination. According to the regulations to the Act, which we translated last year, each of those declarants must provide proof that his or her right to stay in the foreign country has been terminated. This means a cancelled foreign passport or permanent residence card.

The South Korean government data on “reverse migrants” is presented below in tabular form, side-by-side with IRS Federal Register data. “Year” refers to the year to which the Whitebook statistics pertain; the links will bring you to the appropriate section or page (except for 1997 and 1999, where the link goes to a ZIP archive on the MFA website; the relevant file in the archive is #9). Most of these are in PDF, but some are Microsoft Word DOCs or Hangul Word Processor HWP files. If you can’t open the HWP files on your computer, try using ThinkFree, a Google Docs-like online viewer. “Reverse migrants” (역이주, 逆移住) comes from the Whitebook. The Federal Register numbers are my own count, with obvious duplicates eliminated.

Data up to year 2000

In this table, the “ratios” are the ratio between the Whitebook reverse migrant number and the count of names (all names or just Korean-language names as appropriate to the column). A higher ratio means a higher number of people who might be expected to appear in the Federal Register, but failed to do so for some reason — they weren’t long-term green card holders, they forged the proof of abandonment of status that the South Korean government demanded, or the IRS accidentally or deliberately excluded them from their list even though they should have appeared. In 1997, the first year for which Federal Register data is available, the names published include ex-citizens from 1995 and 1996 as well, and so the ratio is calculated based on the sum of the Whitebook data for all three of the years in question.

There are two obvious potential sources of error in my count of Korean names in the Federal Register, which go in opposite directions: the count fails to include Koreans with names that were not obviously Korean (e.g. people with Anglo given names, no middle names listed, and ambiguous surnames such as Lee or non-Korean surnames such as married names), and it includes all people with Korean names even though they may not have gone to South Korea but instead migrated to a third country such as Canada and naturalised there. Since the Federal Register provides no data about people besides their names, it’s difficult to measure the size of either effect, but we can hope that they roughly cancel each other out.

Data from 2001 onwards

Columns in this table are the same as above, with two additions. The first is “I-407” at the right side of the table, the number of people recorded by U.S. Citizenship & Immigration Services as abandoning lawful permanent resident status (i.e. green cards) from 1 June of the previous year to 31 May of the captioned year. We owe many thanks to Shadow Raider for filing a Freedom of Information Request with USCIS to get this data; he scanned it and uploaded it to Google Docs. “Proportion” is the ratio of South Korean declarations of permanent return to I-407 filers, as a percentage. Keep in mind when looking at these column, however, that an I-407 is not the only way to lose a green card; you might also have it cancelled at the airport at the whim of a Homeland Securitate officer who doesn’t like the fact that you travel outside the U.S. too often, as immigration lawyers warn.

The second addition is “Never left”, at the left side of the table. Starting in 2001, the Ministry of Foreign Affairs began dividing “reverse migrants” into those who actually left the country and those who abandoned their plans to emigrate prior to leaving (해외이주 신고후 출국전에 이민을 포기한 사람). These two figures appear to have been lumped together in previous years, meaning data from before and after 2001 are not directly comparable. Up until 2007, no breakdown was provided by country for those who abandoned their plans to emigrate; I have given the total number (but not the by-country number) in the table for reference purposes, but it is excluded from the calculation of the ratios given in the Federal Register column. You’ll note that data for 2002 and 2003 is in italics; for some reason, those years they went back to the old practice of giving the country breakdown with the “never left” number mixed in, meaning those statistics are comparable to the old numbers rather than the new numbers.

Green-card holders

Under 26 USC § 6039G(d), the Federal Register list is supposed to include some ex-green-card-holders: those “long term former permanent residents” who held their green cards for eight or more of the last fifteen years. The USCIS data does not give any indication of how many ex-green-card-holders fall into that category, but the IRS has stated in Paperwork Reduction Act filings in the late 1990s and early 2000s that it expected about 11,000 people to file Form 8854 each year. Take away about a thousand to account for the number of people giving up citizenship in those years, and it seems that the IRS expects that about 10,000 out of the 15,000 to 20,000 people (i.e. 50% to 66%) who give up green cards each year are “long term” holders.

The first thing that’s noticeable from the table of reverse migration data since 2000 is that the USCIS data on green card abandonments seems pretty sensible: in all years up to 2011 — with the exception of 2002 and 2003, when the South Korean number of reverse migrants was quoted with “never left” migrants included — the number of South Korean reverse migrants was between nine and thirteen percent of the total number of U.S. green card cancellations. Since South Koreans make up about 3% of the foreign-born population of the U.S., that suggests as a back-of-the-envelope measure that South Koreans have about three or four times the “reverse migration propensity” as the average immigrant.

This makes sense from a purely practical standpoint: South Korea is the second-richest country among the U.S.’ top 20 sources of new green card holders (the richest is Canada). The next runner-up and the sole other country classified as “high income” by the World Bank, Mexico, has less than half their PPP-adjusted GNI per capita, while the remainder comprise seven upper-middle income countries (China, Colombia, Dominican Republic, El Salvador, Iran, Jamaica, Philippines) and ten lower-middle income countries (Bangladesh, Burma, Cuba, Ethiopia, Haiti, India, Iraq, Nigeria, Pakistan, Vietnam).

So we’re left with the mystery of why the Federal Register data is so incomplete. A Government Accountability Office report confirms that the Federal Register list did not include ex-green-card-holders in 2000. We’ve previous reported that other South Korean former long-term green card holders, such as celebrities Eric Moon & Tony Ahn in 2008 and Ok Taec-yeon in 2010 did not appear either. However, in 2012, the Federal Register list began including a notification explicitly claiming that it did include ex-green-card-holders. No such entry has been confirmed yet.

If the IRS actually did began including long-term green card holders in their list after having not included them before, we’d expect to see at least several thousand additional names in the Federal Register starting in 2012 (based on the IRS’ own apparent estimates of the number of “long term” green-card-holders abandoning status) … but instead the numbers dropped. Even the recent rebounds observed in 2013’s first and second quarters aren’t enough to account for the FBI’s figures on renunciations of citizenship, let alone the expected numbers of ex-green-card-holders.

Citizens

Some of the South Korean reverse migrants are people who gave up U.S. citizenship instead of green cards. According to the Migration Policy Institute, South Koreans have a higher U.S. naturalization rate than average: 54.7% of foreign-born Korean Americans have become U.S. citizens, against just 43% of the overall foreign-born population. Of course, people who actually went to the trouble of naturalising would probably show a lower propensity towards reverse migration, but some people renouncing U.S. citizenship in favour of South Korean citizenship aren’t naturalised immigrants but instead native-born citizens, like these five young men who renounced to serve in the South Korean military (among whom only two showed up in the Federal Register).

However, there’s no reason to expect the proportions of green card holders and citizens among reverse migrants to fluctuate wildly from year to year, nor the proportions of short-term green card holders to long-term green card holders.

So with those two facts in mind, we would expect the ratio between the MFA and IRS numbers to be roughly stable. In reality, we see it rising sharply into the triple digits in 2006 — a year in which other evidence also suggests that the IRS for some unknown reason began to exclude many names from their “name-and-shame list”, resulting in an artificially deflated count of ex-U.S. citizens. The MFA-to-IRS ratio remained high until falling back down to historical levels in 2009. It jumped in 2010 & 2011, but surprisingly it fell again in 2012, a period for which we otherwise have have very good reason to believe that the Federal Register drastically under-reported the number of ex-citizens.

I can’t think of any factors within the IRS which would cause the rate of non-publication of the names of South Korean ex-U.S. citizens to fall, but the overall rate of non-publication to rise. Perhaps some of the errors we’ve been blaming on the IRS are instead attributable to the State Department or the individual consulates failing to forward Certificates of Loss of Nationality? But then State seem to be communicating fine with the FBI, which uses those same CLNs to maintain its gun control database. As with every government data release, we have more questions than answers.

Falling number of reverse migrants from 1994 to 2008

One other curious thing to note is that the number of reverse migrants reported has fallen sharply over the years. It has bounced back quite a bit recently — a fact also noted in anecdotal reports — but still hasn’t reached the levels seen in the 1990s, when South Korea was a much poorer country. A number of factors could contribute to this trend:

The cancellation effect: The 2001 separation of the MFA data into actual returning emigrants and those who abandoned plans to emigrate caused a far larger drop in the number of US reverse migrants than non-US reverse migrants. The obvious interpretation is that many of the US reverse migrants recorded in 2000 and earlier never actually went to the US; apparently they got green cards but then decided not to take advantage of them for whatever reason. The falling number of people who abandon their plans to emigrate in recent years may offer further support for this theory: it may indicate a lower number of people who get in line for family reunification green cards but vacillate about actually using them when they’re finally approved.

The Sa-Yi-Ku effect: Many South Korean immigrant shopkeepers in Los Angeles lost their livelihoods in the riots which resulted from the acquital of the four police officers who assaulted Rodney King. Some sources claim this resulted in large amounts of reverse migration. However, interviews on the ground don’t seem to confirm this. Nancy Abelmann & John Lie found quite a bit of resistance to the idea of reverse migration in their 1995 work Blue Dreams:

(page 17) “If people think that there is a distance between blacks and Koreans, it is nothing compared to the distance between Koreans here [United States] and there [South Korea]. If they think blacks blew up at Koreans in South Central, watch what would happen if all of them [Korean Americans] went back to Korea today.” He suggested that Korean immigrants are by now unwelcome in South Korea. “The Koreans in the United States,” he stressed, “are all people who have been looked down upon in Korea.” Failures at home, Lim’s immigrants arrive in the United States with their eyes still fixed on South Korean compatriots and cohort groups, from whom they seek redemption and clamor for respect … “there is no one who knows they can make it in Korea who comes [to the United States]” …

(page 36) Yu and two other female stall managers who gathered in the tiny space of her stall — some four feet square — all proclaimed that they hated the United States, that they wanted to go back to Soth Korea, but they could not: “Even if you come here with lots of money, it all gets tied up in payments — house, car, and so on — it’s different from Korea, you can’t leave.” … A Korean American mental health counselor has commented that the “second stage” of the riot response, after the immediate shock and anger of loss, is to regret having come to the United States in the first place …

If there were a lot of reverse migrants due to the riots, it might explain both the high numbers of reverse migrants reported by the South Korean government in the early-to-mid 1990s; the high number of Korean names in the 1997 Federal Register lists (which, as mentioned, also cover 1995 and 1996) may represent the tail-end of this effect. However, it wouldn’t explain why reverse migration from other countries such as Canada & Argentina also decreased in tandem; the latter is probably due to the effects of the 1997 Asian financial crisis on South Korea.

The visa effect: as the economic gap between South Korea and the United States narrowed, the U.S. Embassy in Seoul has become far less likely to deny non-immigrant visas to South Koreans on the grounds of alleged immigrant intent; thus, South Koreans who in past years used family preference green cards as a sure (if slow) way to study & work in the U.S. for a few years may simply instead be obtaining the appropriate F, H, or L visas, so that they make neither declarations of emigration when they leave nor declarations of permanent return when they come back.

The lawbreaker effect: The rate of compliance with the Overseas Migration Act may have dropped. This probably takes the form of people emigrating in earlier years without reporting it, meaning that they avoid the need to make a declaration of permanent return when they come back; it’s less likely that registered emigrants are sneaking back to the country without making declarations of permanent return. A certain rate of non-compliance is to be expected all along; a civilised country does not impose life-altering fines on emigrants for failure to comply with such obscure paperwork, the way the U.S. does.

Conclusion

The number of emigrants and renunciants itself may be small compared to the number of new immigrants to the U.S., but the problems with the data bring into stark relief the IRS’ level of competence on matters relating to U.S. Persons abroad: if they can’t even do something so simple as publish a complete list of names, how well do you think they’re handling more complicated issues, such as the fair treatment of people who don’t actually owe U.S. tax but have fallen behind on useless paperwork which threatens tens of thousands of dollars of fines?