Double Counted in the Census

Imagine yourself as an immigrant to America in the early 20th century.  You are happy with your decision to leave your homeland for a new life in America. Perhaps after a few years you saved enough money to send for your wife and children to join you. You have found a job, and you have found a house to live in. Perhaps you don’t yet understand the English language perfectly yet, but you are slowly learning. You may not get much practice with English though, because  your neighbors and co-workers speak your native language. One day someone knocks on your door – they are from the government, and they ask all sorts of official questions. “Who lives here?” “What are the names of your family members?” The questions were dutifully answered.

Fast forward eighty or one hundred years. Descendents of those immigrants pour over online or microfilmed images in search of answers about their ancestors. Families are found! But…is the information correct? Most of the time, it is correct. But not always, at least not in my family. Ignoring the numerous name spelling errors, the most unusual census mistakes in my family involve relatives that were counted twice!

All My Children

The first example of this was in the 1910 census for the family of Joseph and Antonina Pater (which is listed as “Potter”, or how Pater sounds in Polish).  In 1910, most of the family was living just outside of Philadelphia in the Bucks County borough of Attleboro (today known as Langhorne). Because Antonina’s mother had recently arrived and she was the oldest family member, she is listed as the head of the household (F. Annie Pluta indexed as F. Amie Theta…seriously, it’s a wonder I find anyone in the census!).  The 70-year-old F. Annie is followed by Joseph and Antonina and their six children (although there is some confusion as some are listed as grandchildren of the head of the household and others as children). The only problem? The two eldest daughters, Frances and Eva (listed as Francesca and Edna), were already married with children and living elsewhere.

Frances’ husband Paul and their son Edmund may be enumerated as a separate family underneath the Pater clan (listed under equally mangled and hard-to-read names). Eva, her husband Edward Süsser, and their children Edward and Anna are all enumerated on the census in Dover, Morris County, New Jersey (as the “Züsser” family).  In this case, only Eva is counted twice since I did not find another listing for Frances.  I believe that the married children who were not actually living with their parents were listed simply due to a language mis-understanding when the census taker asked for the names of their children.  By 1920, the Pater parents only list those children still living with them (Walter and Victoria).

By Any Other Name

A more curious case of double-counting happened in the 1930 census.  My Piontkowski ancestors, John and Rose, had been living in the United States for 25 years, so I would have assumed they had a better understanding of both the English language and what the census-taker wanted after having participated in two other federal censuses. The couple leaves out their daughter, who by this time had married and left the family, but counts their teenaged son, James, as well as their married son Joseph, his wife Catherine, and their daughter, Josephine. The entire family lives on N. Front Street in Philadelphia.

I knew that Joseph Piontkowski later used the surname Perk, but I never thought to look for Joseph Perk on the census.  Why should I? I had already found him living with his parents.  Only he really wasn’t living with his parents in 1930! I recently got in touch with my cousin, Joseph’s daughter, who had been researching her family.  When she wrote that she found the Perk family listed in the 1930 census, I did a double-take.  Sure enough, they are living on Hancock Street in Philadelphia about a mile away from his parents.  Listed are Joseph Perk, wife Katherine, daughter Josephine, and daughter Jean – who, based on the age of 0/12, had just been born!  Anyone without knowledge of the name change would certainly think that these were two different families, but they are the same.

I wonder how inflated the census numbers are/were due to difficulties with immigrants understanding the questions? Oh well, Eva Süsser, Joseph, Katherine, and Josephine Perk may all have been counted twice in one census or another – but at least that makes up for my grandmother Margaret Bergmeister not having been counted at all in both the 1920 and 1930 census!


Census Records: What’s in a Name?

When I first got started in genealogy, I thought the Soundex was an amazing thing. It helped me find many incorrectly written names, often simply mis-pronounced by the foreign speaker or mis-understood by the American census taker. But, the Soundex only gets you so far…some errors are just too much to overcome. For example, the Soundex assumes that the first letter of the surname is correct, but what if it’s not? Thanks to computers and indexing, finding someone on the census is a lot easier than it used to be.

Zawodny Census Names

An example of a family that was hard to locate in the census is my Zawodny ancestors. As Polish surnames go, the name of Zawodny isn’t all that hard or unreasonable! But if you try to find them in census records, good luck. You’ll find three different names! The family first arrived in 1902, so the first census year is 1910. This is how the family’s entry compares for 1910, 1920, and 1930:

1910 – Savonia, Joseph, age 28. Wife Mary, age 20.

1920 – Cawodny, Joseph, age 39. Wife Laura, age 36.

1930 – Zavodny, Joseph, age 50. “Sister” Laura, age 44.

As you can see, only the 1930 surname would have been found using a Soundex search. The wife’s name changes, most likely because her Polish first name Wacława doesn’t really translate into an English name, at least not the same way that Jozef becomes Joseph.

Another favorite family in census records is my Piontkowski ancestors. While the 1920 entry of “Pontdowke” and 1930’s “Peontkowski” show up in the Soundex, the family’s whereabouts in 1910 had me stumped. Finally, I found them – listed under “Kilkuskie”.  Not really an intuitive search, but the first names, ages, neighborhood, and other information all matched. The best part about their entries are the ages – while the husband’s age is or at least close to what it actually was for those census years, or ages 39-49-59, the wife seems to grow younger each decade. Perhaps it was unfashionable back then for a wife to be five years older than her husband, but her ages show up as 37-52-54 while her actual age was 44-54-64!

So, how do you find someone when the surname isn’t right and Soundex searches fail you? The old standby prior to computers was to search for the known address. In the case of these two families, they each had a different address for each census year. If a family moved frequently, even though they stayed in the same neighborhood, they’ll be difficult to find unless you know through some other means, such as a city directory, what their actual address was during the census year.

One method that I used to find these records when “last name” searches failed was to search with a combination of the first name, approximate age, and country of birth. It helps if you know at least the county or city where the family lived, because you may get over a hundred men named “Joseph” born in “Poland” or “Russia” around 1879. But, by carefully checking the other family members, you will find the family if they are there. You can also combine a search using these elements with a spouse’s first name, or a parent’s first name if you are searching for a child. Try using Steve Morse’s searches if other search sites have you stumped.

This post is an excerpt from a future article on Searching US Census Records.