The 1940 Singh Census, and the attack on Indian immigrants

A new history dataset shows the impact of immigration restrictions

I’ve been analyzing U.S. Census data to understand the history of Sikh and other South Asian Americans, and I’m excited to release a public dataset of almost a thousand people with the last name “Singh” living in the United States in 1940.

(The source data is available on Github in JSON format. Download it now!)

This census data gives us a snapshot of life during the period when immigration from Asia was largely shut down as a result of the Immigration Act of 1924. It wouldn’t start back up until 1946, when the Luce-Cellar Act would allow for 100 new immigrants from India every year.

1. Where were India-born Singhs living?

I took every Singh born in India, took their census location and count, and ran it through OpenHeatMap. The results are pretty clear: early Singhs were very heavily clustered in California.

2. How old were India-born Singhs in 1940?

The population of Singhs in 1940 was much older than I’d expected:

• 1940: the median age of India-born Singhs was 51
• 2016: the median age of India-born people was 38.5

The lack of younger immigrants is a clear marker of the immigration restrictions put into place starting in 1924, which cut off the flow of younger immigrants.

3. Where were married women named Singh born?

Anti-Asian immigration restrictions left South Asian Americans with a severe gender gap.

This led many Sikh Americans to marry women of Mexican descent, escaping anti-miscegenation laws, as documented in books like Karen Leonard’s Making Ethnic Choices.

We can see footprints of the Punjabi-Mexican marriages in the data. Out of the people with the last name “Singh” who are designated as wives, over half were born either Mexico, or in U.S. states that were formerly part of Mexico.

Here’s the breakdown by state/country, where 🇲🇽 = current/former Mexico:

1. Mexico (40) 🇲🇽
2. Other US states (14)
3. California (10) 🇲🇽
4. Texas (9) 🇲🇽
5. New Mexico (7) 🇲🇽
6. New York (5)
7. Illinois (2)
8. Arizona (2) 🇲🇽
9. India (2)
10. British South America (1)
11. England (1)
12. Philippines (1)
13. Spain (1)
14. West Indies (1)

4. What were the most common names of the children of Singhs?

I counted the names of every child of a named Singh — doing my best to remove duplicated children, when multiple adults (e.g. husband and wife) in the same district have a set of children with exactly the same names.

All of the names are Anglo and/or Hispanic. While I don’t know the reasons behind this, I’d guess this might be an assimilation strategy during a time of increased racism, as well as an indication of the important role that non-Sikh women played in family life in a new country.

1. Mary (8)
2. Carmen (6)
3. Harry (5)
4. John (5)
5. Gloria (4)
6. Peter (4)
7. Raymond (4)
8. Robert (4)
9. Albert (3)
10. Carmelita (3)

5. How were Singhs racially categorized?

Most people with the last name “Singh” were categorized as “Hindoo” (i.e. South Asian), rather than “White.”

But racial perceptions were highly changeable. For example, people may have been perceived differently, based on immigrant generation and marital status.

Looking at male Singhs, ages 18 and up:

• single men born in India were overwhelmingly categorized as “Hindoo” (i.e. South Asian)
• married men born in the United States were more frequently categorized as “White”

Were those Singh men in closest proximity to Anglo or Mexican spouses/communities most likely to be seen as White, i.e. less foreign?

Why this matters

By 1940, the American Singh community had borne the scars of two decades of racist immigration restrictions can shape a community. Looking at this census data can help us understand those historical impact, and can be analyzed alongside works like Karen Leonard’s Making Ethnic Choices: California’s Punjabi Mexican Americans, which helps us understand the lived meaning behind the data patterns.

The 1940 Singh Census is also a reminder to document and measure the impacts of racist and xenophobic attacks today, so we can zoom out from individual incidents to see the whole. Whether we’re documenting anti-trans bills, or measuring changes in immigration applications immediately after the Trump election and Muslim Ban, understanding the patterns are a key part of our advocacy and resistance.