How the 1965 Immigration Act Quietly Rewrote Who Comes to America

Name the one federal law, signed in a ceremony on a small island in New York harbor, that did more than any other to decide who the next generation of Americans would be. If you said the Civil Rights Act, you are in good company and you are wrong. If you said the Voting Rights Act, which was signed that same summer, you are closer to the right neighborhood and still wrong. The answer is a bill most Americans could not name if you gave them a hint and a lifeline: the Immigration and Nationality Act Amendments of 1965, known by the names of its House sponsor Emanuel Celler of New York and its Senate sponsor Philip Hart of Michigan. Hart-Celler. Lyndon Johnson signed it on October 3, 1965, at the foot of the Statue of Liberty on Liberty Island, with Ellis Island in the background, and the stagecraft was no accident.

I'd argue that the bill was right about one big thing and careless about another, and the American people have been living with the difference ever since. Let me take them in order.

What the law was fixing

To understand 1965 you have to understand 1924, which is told in The 1924 Law That Built the Quota System and the Border Patrol. That law handed out visas by national origin: a country got a quota tied to the ancestry of the American population, and the biggest allotments went to Britain, Ireland and Germany, while Italy, Poland and Greece got slivers and most of Asia got nothing. Forty years later nobody could defend that with a straight face. The country had just fought a war against a regime that ranked people by blood. It had passed the Civil Rights Act in 1964. A Congress that told Black Americans their citizenship was equal was hardly going to tell Italian Americans that their cousins were second-rate.

So the central purpose of Hart-Celler was good, and conservatives ought to say so. It abolished the national-origins quota system, phased out over a few years. Under it, you could no longer be turned away because of what country your grandfather came from. That is consistent with the Declaration's claim that all men are created equal, and consistent with the old Republican view, going back to Lincoln, that the American creed belongs to anyone who accepts it.

The question is not whether to end the quotas. The question is what was built in their place.

What the law built

Congress did not replace one sorting rule with a blank slate. It replaced it with a new rule, and the new rule was family.

Here is the architecture, in plain terms. Three things sat in the 1965 law.

One. Immediate relatives of U.S. citizens, meaning spouses, minor children and parents, were placed outside the numerical limits altogether. There is no cap on them in the law, which means the total in any year depends on how many citizens have such relatives and how many choose to sponsor them.

Two. Seven preference categories divided the capped numbers. In the order of the original statute they were: unmarried adult sons and daughters of citizens (20 percent of the visas), spouses and unmarried children of permanent residents (20 percent), professionals, scientists and artists of exceptional ability (10 percent), married sons and daughters of citizens (10 percent), brothers and sisters of citizens (24 percent), skilled and unskilled workers in occupations where labor was short (10 percent), and refugees (6 percent). Add them up and you get 20 + 20 + 10 + 10 + 24 + 10 + 6 = 100. Now add the four family categories: 20 + 20 + 10 + 24 = 74. Seventy-four percent of the capped visas went to relatives. Only 10 percent and 10 percent went to professionals and to workers, and even those two required a Labor Department certification that American workers were not available.

Three. Numerical ceilings and a per-country limit. The Eastern Hemisphere got a ceiling of 170,000 a year, with no more than 20,000 from any one country. The Western Hemisphere, which had never had a numerical limit, got one for the first time: 120,000 a year, starting in 1968. Add those: 170,000 + 120,000 = 290,000. That 290,000 is the commonly cited headline number for the preference system, remembering that immediate relatives fell outside it.

The Western Hemisphere cap is the part that rarely gets told. Restrictionist members of Congress insisted on it as the price of ending the old quotas. They feared that, with the old system gone, the country's neighbors would become the main sending region. The cap was a compromise, and it had a consequence nobody in 1965 spent much time on: a border that had been governed by labor programs and informal movement was now governed by a visa quota too. For decades afterward, the gap between the quota and the demand was filled by people who came outside the system. You can see where that leads in Reagan Signed the 1986 Amnesty and Got Only Half of the Bargain.

The reassurances, and what followed

Supporters in 1965 told the public and their colleagues that the bill would not change the character of the country. They said it would not flood the nation and that the ethnic mix would not be turned upside down. That was the pitch, and it was repeated in the Senate hearings and in the floor debate. Check the committee record for the exact language before you quote anyone, because the specifics matter and this subject draws plenty of misquotation from every side.

What followed was a major change. The share of arrivals from Europe fell. The share from Asia and Latin America rose. The reason was not mysterious. The new system rewarded family ties, and family ties spread outward in a chain: once a person is a citizen or permanent resident, he can sponsor a spouse and children, and later siblings and married children, each of whom, once settled, can sponsor their own. Critics call this chain migration. Supporters call it family reunification. Both are describing the same mechanism, and the choice of word tells you which side of the argument you are on.

I do not think the people who wrote the 1965 law were lying when they said they expected modest change. I think they believed the old European-heavy applicant pools would keep using the new system. They did not model how family chains work over decades. That is a design failure, not a conspiracy, and it is the right frame for the conservative argument that follows.

The strongest case for the 1965 design

Let me state the opposing view at its best, the way a defender would put it.

One. Family is the most conservative institution there is. A system that lets Americans bring their spouses and children, and in the original design their brothers and sisters, strengthens families.

Two. Family-sponsored immigrants come with a built-in support network: a relative who vouches for them, helps them find a job and a place to live, and signs a legal affidavit of financial support. That lowers the burden on public programs.

Three. The old system of ranking nationalities by ancestry was worse. Replacing it with a neutral rule was a moral gain.

Four. Skilled and educated immigrants have always come in through the family system too. The brother of an engineer may be an engineer.

Those points are not nonsense. Point one has real force for spouses and minor children, and nearly every conservative reform proposal keeps those. Point three is correct. If you want to argue with 1965 you have to concede that the system it ended was indefensible.

Where the design fails

Now the case against the design, in numbered premises.

One. It sets the size of immigration by who already lives here, not by what the country needs. Under the family system, the number and the character of the next wave is decided by the relatives of the last wave. A citizen's decision to petition for an adult brother is a private choice with a public consequence. There is nothing wrong with the private motive, but a country that sets its numbers entirely through private motives has given up making the choice itself.

Two. The preference for skills was an afterthought. In the original 1965 structure, workers and professionals got a fraction of the capped numbers, while family categories took the bulk. Compare that to the systems of other advanced countries, which usually assign a bigger share to work and skills. A country that wants its immigrants to fill needs, start businesses and raise the productivity of its economy has every right to weigh those things.

Three. The per-country limit makes the lines absurd. The 1965 law capped the Eastern Hemisphere at 20,000 per country, and later Congresses extended a per-country limit to the whole world. That sounds fair. The result is that people from countries with large numbers of applicants wait for years, even decades, in some categories, while people from smaller countries move up quickly. A rule meant to prevent one nation from dominating the flow ends up punishing the family of an applicant for being from a place many others want to leave. See The Four Doors Into America and Who Holds the Keys to Each for how the doors work today.

Four. A queue measured in decades is not a queue. If a sibling category keeps people waiting longer than a working lifetime, the promise on paper is no promise at all. It invites people to come another way, which is the opposite of what an orderly system is for.

Conclusion: the design should be rebuilt around the idea that the nation chooses its numbers deliberately. That is not a radical idea. It is how the rest of the law is supposed to work.

What a conservative fix looks like

None of this requires re-fighting the moral question about 1924. Keep the end of the national-origins system. Argue about what replaces the sorting. Here are the main proposals that have circulated among Republicans, with the caveat that conservatives do not all agree on them.

• Keep the nuclear family. Spouses and minor children of citizens and permanent residents stay at the front of the line.
• Narrow the extended-family categories. Adult siblings and married adult children are the categories most often named for reduction or elimination. A bill in the Senate in 2017, the RAISE Act, offered by Senators Tom Cotton and David Perdue, would have cut the family categories back to the nuclear core. It did not become law. Read its text on Congress.gov rather than relying on anyone's summary, including mine.
• Shift weight to work and skills. Raise the share of visas that go to people who bring needed skills, English and a job offer. The mechanics of the current employment categories are in H-1B, H-2A and H-2B, What Each Work Visa Does and Who Can Use It.
• Reform the per-country limit so that it does not leave applicants from the biggest source countries behind a wall of decades.
• Keep refugees and asylum honest and separate from the family numbers. See Asylum, Refugee, Parole and TPS, Four Statuses People Mix Up.

I would add a principle: whatever you pick, pick it by vote. The point of reform is to move the decision from the accidents of family chains to the deliberate judgment of the elected Congress. That is conservative in the plainest sense. The Constitution gives the naturalization power to Congress, not to a pipeline.

Don't relitigate motives

A word about tone, because this subject invites bad behavior on both sides.

Some critics of the 1965 law claim it was a plot to change the country's makeup. I do not think the evidence supports a plot, and I do not think conservatives who want reform need that theory. Some defenders claim that anyone who wants to reform the family categories is secretly trying to bring back the 1924 quotas. That is just as unfair. Both claims treat a debate about design as a debate about hearts.

The honest argument is narrower and stronger. We can admit that ending ancestry quotas was right. We can admit that supporters' predictions in 1965 were wrong in ways nobody fully foresaw. And we can propose a different design for the next sixty years without insulting anyone's grandparents. If you want to practice saying that out loud, How to Talk About Immigration at the Table Without Losing the Room is the post to read next, and The Case for Being Pro-Immigrant and Pro-Border at the Same Time makes the broader argument.

It also helps to see the 1965 act as one chapter in a long story. The 1986 act, the 1990 act and the 1996 law each tried to patch part of the structure, and Four Decades of Republican Immigration Fights, From Reagan to the Gang of Eight shows how often the party has divided over the patches. The 1965 foundation is still there under all of them.

The bottom line

• Hart-Celler, signed October 3, 1965, was right to end national-origin quotas. A conservative should say so without hedging.
• It replaced them with a family-based system in which about 74 percent of the capped visas (20 + 20 + 10 + 24 = 74) went to relatives and about 20 percent to workers, so the size and shape of immigration became a function of who was already here.
• The fix is to keep spouses and minor children first, weigh skills and work more heavily, and rework the per-country limits, and to do it by vote in Congress.
• Argue the design, not the motives. Nobody in 1965 needs to be a villain for the design to need repair.
• When someone quotes a 1965 promise or a 1965 number at you, ask for the page. The hearings and the statute are on Congress.gov and in the National Archives.