A Diaspora Is One Population Whose Outcome Differs in Every State It Reached
A diaspora is a population that traces itself to one origin place and lives, at the same time, inside more than one destination state, where one origin produces a different legal status, a different set of institutions and a different everyday language in each place it reached.
Three conditions have to hold before such a population can be studied as one, of which the third decides whether any figure about it exists.
Dispersal Across Several Destinations
The population lives in more than one state at once, so what happened to it can be compared across destinations that differ in law, in labour market and in majority language.
An Origin That Both Sides Recognize
Members and outsiders agree on a place the population came from, which is what lets an origin government, a host government or a researcher assign anybody to it.
An Instrument That Records Membership
Somebody keeps a register, a census category, a legal status or an association roll, so that the population has a number attached to it rather than only a story.
The influence of a diaspora on cultural diversity and on identity is confined to the higher level extension, which a standard level course never reaches, though that same course meets diaspora growth as one factor shaping future international tourism wherever the leisure, tourism and sport option is studied.
Common Mistake
A migrant stock is a count of people enumerated somewhere other than the state they were born in, taken on a single date.
A diaspora count reaches descendants, because membership in it passes down a family rather than being fixed by a birthplace.
The two figures therefore separate as a settlement ages, with the gap widening fastest where the destination-born generation is large.
Treating one as a substitute for the other makes a settled community look like a recent arrival.
Three Registers Count One Diaspora and No Two of Them Count the Same People
Any diaspora is recorded three times over, by the state it came from, by each state it went to, and by whoever assembles national tables into a world total, with each of those three registers built for a purpose of its own.
The Origin State's Register
An origin state counts people it claims, so its list can reach citizens of other countries who have never set foot in it.
The Destination State's Instrument
A destination state records residents on whatever variable its own statistical law permits, which may be race, ethnic group, ancestry, citizenship, country of birth, or the language used at home.
The Harmonized International Estimate
A harmonized international estimate can only assemble what some state has already published, so a population no state records separately is invisible in the world table.
Note
A destination recording citizenship and a destination recording ancestry will describe one family differently on the same day.
The origin state's own list can exceed both, because it reaches grandchildren who carry another passport.
Which instrument produced a figure decides which people its total contains.
Setting two of them side by side needs a sentence naming what each one asked before any comparison is drawn.
What a Diaspora Changes in the Place It Arrives Can Be Dated and Counted
A diaspora leaves three kinds of record in a destination, each of which carries a date that a description of atmosphere does not.
Institutions With a Street Address
A school, a place of worship, a mutual aid association, a newspaper or a burial ground is founded on a date and under a legal form, so it fixes when the settlement became permanent.
The Language Used in the Household
Several national censuses ask which language a household uses at home, which turns retention into a number that can be taken twice and compared.
A Legal Category Created for the Population
Where a destination writes a status into law for one arriving group, membership becomes an administrative fact published every year.
Hint
The fabric of the commercial street records the same arrival in signage and shopfronts, which is surveyed by counting features along a fixed transect.
Date an institution before using it as evidence, because a congregation meeting in a rented hall and one that built a hall are twenty years apart.
Founding dates sit in company registers, charity rolls, planning files and the institution's own anniversary material.
An institution that closed is evidence too, since its closing date marks when the population it served moved or stopped needing it.
Retention Is a Comparison Between Generations, So One Census Year Cannot Show It
Retention is a claim about what the generation born in the destination kept from the generation that arrived, which means it needs two groups or two dates.
A census that asks both country of birth and language of the home supplies both halves at once, because it splits the population into arrivals and descendants and asks each the same language question.
The share reporting only the majority language rises with each generation unless something keeps the older language in daily use.
What keeps it there is usually a school, a form of worship or a legal requirement, since those three put the language in a setting where children acquire it rather than hear about it.
Cultural retention and cultural identity therefore part company, because a family can report a strong attachment to an origin in the same census in which it reports speaking only the majority language.
Common Mistake
A language-at-home question records which language the household works in, while the ability to speak the other one goes unrecorded.
A bilingual household that answers with the majority language disappears from the retention figure while keeping the skill.
The question is still worth asking, because asked the same way twice it measures a shift in what gets used.
Read any retention figure as a statement about households, so the claim stays inside what the instrument can support.
Influence Runs Back to the Origin Place, Which Builds Instruments to Reach It
An origin state acquires a population it can claim and cannot govern, so it writes law, creates a body and funds programmes to keep hold of it.
A Definition Written Into Statute
The origin state decides who counts, so the line it draws decides who may enter, work, study or receive its services.
A Named Agency
A body inside the origin government compiles the count and runs the programmes, so the diaspora gains an office and a budget line at home.
A Rule About Voting
Whether members vote in origin elections usually turns on citizenship, which cuts across the descent line the same state used to count them.
Teachers, Curriculum and Schools
The origin state can place its own schooling inside destination cities, which moves a curriculum rather than a person.
Note
The definition is where the politics of a diaspora policy sits, because everything else follows from who is inside it.
A wide definition reaching distant descendants creates obligations the origin state has to fund.
A narrow one reaching only recent emigrants leaves out the communities with the longest history abroad.
Watch for the date a definition changed, since a jump in a diaspora total often marks a redrafted statute.
What members send home in money is a flow recorded on the balance of payments and set against the receiving economy, so the evidence for it sits in financial statistics.
Korea's Seven Million Are Counted by the State They Left and Again by the States They Reached
Seoul Counted 7,006,703 Koreans in 181 Countries at the End of 2024 and Publishes the Figure as an Estimate
The Overseas Koreans Agency reported 7,006,703 people in 181 countries as of 31 December 2024, in the 2025 edition of its Overseas Koreans Status, the count it publishes every two years.
The figure is assembled from returns made by Korea's own embassies, consulates general and branch offices, each reporting for the country it sits in.
Those missions build their returns from the host country's population statistics, from surveys run by Korean community associations, from the consular register of nationals abroad and from the mission's own administrative records, including the overseas voter roll.
Against the 7,325,143 the 2021 edition gave for December 2020 and the 7,081,510 the 2023 edition gave for the end of 2022, the total has fallen by 318,440 in four years.
Common Mistake
No government can run a census abroad, so every origin state's diaspora total rests on somebody else's data plus its own consular paperwork.
A community that runs formal associations is easier for a mission to count than one that runs none.
Coverage therefore varies by destination in a way the single world total hides.
A compiled estimate moves when a mission changes its sources as well as when people move.
A falling total is evidence about the count before it is evidence about the population.
Four Destinations Record the Same People on Four Different Instruments
Korea's own list gives the United States 2,557,047, China 1,848,241, Japan 960,970 and, in Central Asia, Uzbekistan 175,338 alongside Kazakhstan 122,554.
Each of those states records the same people on an instrument written for its own purposes, so no two of the instruments ask a comparable question.
Note
The United States asks residents to identify themselves and asks every household which language it uses.
Japan records a legal status created by one treaty and reports how many people hold it.
China records a nationality inside its own citizenry and therefore publishes no diaspora figure at all.
Kazakhstan records a national group among its own citizens, so its census carries a Korean row and no diaspora column at all.
The United States Asks Who People Say They Are and What They Speak at Home
About 2 million people in the United States identified as Korean in 2023, on Census Bureau estimates reported by the Pew Research Center.
Seoul's figure for the same country is 2,557,047, so the two counts of one population differ by roughly half a million.
The gap is what two questions produce, because a household survey asks residents what they are while a consular network reports who has come to its notice.
The American Community Survey also asks which language is used at home, recording 1,075,247 Korean speakers for 2019 among the 67.8 million people aged five and over who spoke a language other than English at home.
Pew puts, for 2023, immigrants at 56 percent of the Korean population of the United States, so a little under half of it was born there.
It also finds 68 percent of Koreans aged five and over speaking English proficiently, a share that includes 39 percent who speak only English at home.
The 1,075,247 speaking Korean at home in 2019 and the 2,557,047 Seoul counted at the end of 2024 were collected five years apart by two governments asking different questions, so neither supplies a retention measure on its own.
Example
The census asks the household and the consulate asks the community, so the two arrive at different totals for one street.
Each figure answers the question it was collected for, because the two questions were written by governments with different purposes.
The 39 percent speaking only English at home is the number that moves with generation, since it rises as the destination-born share grows.
A claim that this diaspora is assimilating has to rest on that share taken twice, at two dates, on the same question.
Japan Records a Legal Status That Can Only Get Smaller
The Immigration Services Agency of Japan recorded 4,125,395 foreign residents at the end of 2025, a rise of 356,418 or 9.5 percent on the year before.
Korean nationals were the third largest group among them, at 407,341, which is 553,629 fewer than the 960,970 Seoul counts in the same country.
Of all foreign residents, 266,896 held the status of special permanent resident, against 274,023 at the end of 2024.
That status comes from Act No. 71 of 10 May 1991, made for people who lost Japanese nationality under the Treaty of Peace with Japan and for their descendants born in Japan.
Article 2 of the Act reaches only those resident in Japan continuously from no later than 2 September 1945, those born there to such a person before the treaty took effect, and their own descendants born in Japan.
Nobody can enter that category from outside Japan, so its only source of new members is a birth to somebody already inside it.
One category therefore lost 7,127 people in a year when the foreign population around it gained 356,418, which is a fall the shape of the law produces on its own.
Common Mistake
Special permanent residency covers people of Taiwanese as well as Korean descent, so the 266,896 is not a Korean total.
What it measures is a status defined by one treaty and closed to new entrants.
A destination that records legal status where another records ancestry produces a count that shrinks as a community naturalizes into the majority.
Reading that shrinkage as departure inverts what the instrument is saying.
China Publishes No Diaspora Count Because Its Instrument Is a List of Nationalities
Seoul counts 1,848,241 people in China, though China's own statistical system holds no category that corresponds to them as a population belonging somewhere else.
China classifies its population into the Han majority and 55 minority nationalities, among which the Korean group, the Chaoxianzu, is one.
The seventh national population census, taken on 1 November 2020, put 1,286,311,334 people or 91.11 percent in the Han group and 125,467,390 or 8.89 percent in the minority nationalities.
Everybody inside that second figure is a Chinese citizen, so the instrument records a domestic minority and has no way to record a diaspora at all.
Where a minority nationality is concentrated, China can constitute an autonomous area for it, which for the Korean group is the prefecture of Yanbian, in Jilin province on the border with North Korea.
The Law on Regional National Autonomy, adopted on 31 May 1984 and amended on 28 February 2001, sets out what such an arrangement must do.
Article 21 requires the organs of self-government to use one or several languages commonly used in the locality, allowing the language of the nationality exercising autonomy to be the main one only where several are in use.
Article 37 requires schools where most students come from minority nationalities to use textbooks in their own languages and those languages as the media of instruction wherever possible.
Whether Korean is used in a Yanbian classroom is therefore settled by an instrument of the state rather than by a household, which is how the Legislative Affairs Commission of the National People's Congress Standing Committee came to report Yanbian's own 2004 regulations on Korean education inconsistent with the national law on the standard spoken and written Chinese language.
Note
A statute can put a language into official and classroom use where a household's own choice would not.
Article 37 qualifies its duty with the words wherever possible, so the medium of instruction stays a decision of the state rather than of the household.
The same origin population in the United States meets no such requirement, where 39 percent of it speaks only English at home.
One origin produced two settings for the same language, because in one destination a statute decides where it is used and in the other a household does.
Central Asia Holds a Community That Was Placed Where It Is by Decree
Seoul counts 175,338 people in Uzbekistan and 122,554 in Kazakhstan, in two republics whose Korean population was created in 1937 by a decision taken in Moscow.
Joint resolution No. 1428-326ss of the Council of People's Commissars and the Central Committee, dated 21 August 1937, ordered the removal of the Korean population from the border regions of the Far Eastern Territory.
The People's Commissariat for Internal Affairs, the NKVD, reported the operation complete on 29 October 1937, with 36,442 families and 171,781 people dispatched in 124 train echelons, most of them to the Kazakh and Uzbek republics.
A population moved by decree chooses no destination, though this one brought institutions with it, because the Korean theatre founded in Vladivostok in 1932, the Korean-language newspaper later renamed Koryo Ilbo and the Korean pedagogical institute were all relocated to Central Asia by the same operation.
Both republics record that population as one of their own national groups, with the Korean one among the ethnic groups the Bureau of National Statistics counted in Kazakhstan's 2021 census.
One community therefore sits in two national statistical systems at once, as a diaspora in Seoul's register and as a domestic nationality in Kazakhstan's.
Theory of Knowledge
Two communities with one origin can differ as much in how they arrived as in where they ended up.
A voluntary settlement decides its own destination, because its first arrivals shape where the later ones go.
A population moved by decree makes neither of those decisions.
Ask what institutions a community was in a position to build, so that the difference between the two routes becomes a question with documents behind it.
Seoul Writes Down Who Counts, and a Court Once Made It Redraw the Line
The Framework Act on Overseas Koreans, Act No. 19402, was promulgated on 9 May 2023 and came into force on 9 November 2023, with its Article 11 in force from 5 June 2023.
Article 2 defines an overseas Korean as a national of the Republic of Korea staying long term or holding permanent residence abroad, or a person who once held that nationality by birth and no longer does, or their lineal descendant.
The Overseas Koreans Agency, an affiliate of the Ministry of Foreign Affairs, compiles the country-by-country count under that definition.
An earlier statute drew that boundary somewhere else, because the Act on the Immigration and Legal Status of Overseas Koreans, Act No. 6015 of 1999, gave rights of entry and of economic activity in Korea to foreign nationals of Korean descent.
As applied, it reached only people who had themselves held the nationality of the Republic of Korea, and their descendants, which left out everybody who had emigrated before the Republic was founded in 1948.
That exclusion covered most of the community in China and almost all of the community in Central Asia, because both movements happened before 1948.
The Constitutional Court held on 29 November 2001, in case 99Hun-Ma494, that the exclusion breached the equality guarantee in Article 11 of the Constitution.
The definition was afterwards widened to take in those who left before the government of the Republic was established, so the two oldest communities entered the category by a court's order at a moment when nothing about them had changed.
Example
One government counts 7,006,703 members of this diaspora while opening its ballot only to those holding its own citizenship.
The National Election Commission opens overseas voting in presidential and proportional National Assembly elections to Korean citizens aged 18 or over.
A Chinese citizen of Korean descent in Yanbian sits inside Seoul's count and outside its electorate.
Membership and franchise are two boundaries drawn by one state, so a diaspora policy has to decide separately where each one goes.
The Origin State Puts Schools Inside Its Destinations, and Their Map Follows Postings
Korean schools abroad are established by a juristic person with the approval of Korea's Minister of Education, under the Act on the Educational Support for Overseas Korean Nationals, which defines them as schools providing Korean school education to nationals of the Republic of Korea residing in foreign countries.
Thirty-four of those schools were operating in 16 countries in April 2026, teaching 12,949 pupils between them.
The United States holds the largest counted population of all and has no Korean school at all, so the network follows where Korean nationals are posted on fixed terms rather than where the counted population is largest.
Hint
Read an institution by who it admits and by what it teaches towards.
A school preparing pupils for examinations in the origin state is an outpost of that state's system.
A weekend class run by parents in a rented room is an institution of the settlement itself.
The two leave different records, because one appears in a ministry's list and the other in a community association's minutes.
Case study
Essay use
Use the Korean diaspora to show how one origin population develops different identities and institutions under contrasting destination-state rules.
Evidence
South Korea estimated 7,006,703 overseas Koreans in 181 countries at the end of 2024, while the United States records ancestry and home language, Japan records a closed legal status, China records Korean nationality within its citizenry, and Kazakhstan records a national group.
Analysis
Each instrument selects a different population, and language retention, citizenship and community institutions then respond to the laws and histories of the destination.
Evaluation
Origin-state estimates cannot be treated as destination censuses, a falling legal-status count may reflect naturalisation rather than departure, and comparisons are valid only when the register, date and question wording are named.
Essay judgement
Overall, the evidence from the Korean diaspora shows how one origin population develops different identities and institutions under contrasting destination-state rules.
Five Questions Produce the Evidence on Any Diaspora, and Most Have Published Answers
For a large diaspora each of the five answers is already published, because the origin state, the destination's statistical office and the institutions themselves all keep records for their own purposes.
Who Counts It, and Under Which Definition
Find the origin state's own register and the statute behind it, because the definition fixes who is inside the total before anybody is counted.
What Each Destination Actually Records
Read each destination's census or residence register for the variable it collects, since that decides what can be said about the population living there.
Which Institutions Have Addresses and Dates
A school, a congregation, an association or a newspaper carries a founding date and a legal form, which dates a settlement more precisely than recollection does.
Tip
The origin state's statute and its diaspora agency are usually on one government website, where the agency's own count is the fastest route to a country list.
A destination's answer sits in its census questionnaire, which most statistical offices publish alongside the tables.
Read the questionnaire before the table, because the wording tells you what the number can and cannot support.
The Last Two Questions Reach the Household and the Origin Government
What the Household Speaks
A census that asks the language of the home gives a retention measure when the question is repeated, which is why no retention claim can be built from a state whose census does not ask it.
What the Origin State Does About It
Look for an agency, an electoral rule and a schooling programme, because those three show what the origin place has decided the population is for.
Exam technique
Name the register before the number, because a diaspora figure without its compiler cannot be set beside another one.
Where two counts of one destination disagree, say what each question asked before deciding whether either of them answers yours.
One destination carried through its law, its count, its institutions and its household language is worth more than five destinations listed.
A claim that a diaspora is losing its culture needs the same trait measured at two dates in the same place.
Where a destination collects no variable on the population at all, say so, because the absence is itself the finding.
What Would Settle a Disagreement About a Diaspora and Its Culture
A disagreement about size is settled by naming the register and the definition, which is why 2,557,047 and 2 million for one country are not in conflict.
A disagreement about retention is settled by one question asked twice in one place, which is what a language-at-home item supplies wherever a census carries it.
A disagreement about influence is settled by a count of something in the destination that was not there before and that carries a date.
Naming which of those three a question is about turns an argument about a diaspora into one with an answer.
Active recall
Which three registers can produce a count of one diaspora?
Why can the number of special permanent residents in Japan fall in a year when the foreign population rises by 9.5 percent?
Which two articles of China's Law on Regional National Autonomy govern the language used by the organs of self-government and the medium of instruction in schools?
Who was excluded by the definition the Constitutional Court struck down in case 99Hun-Ma494?
What does a language-at-home question measure that a membership register cannot?
