“Where are they all kept?”: Undercounted Immigrants in the 1900 Census

When I first started researching my Homestead forebears, I fretted over why I could not find my one-and-a-half year-old grandfather and his family in the 1900 census for Homestead.  I assumed it was because I was a newcomer to genealogy.  I anticipated that, in time, I would gain the skills necessary to find them.  As the years passed, I did become a better researcher, willing to read page-by-page through the census of their neighborhood, but I never did find them.  Eventually, I uncovered the real reason for their absence, hidden in miles of Homestead newspaper microfilm:  their neighborhood had been terribly undercounted, and at the time, everyone knew it.

Representative Homestead newspaper headlines from the 1890s

In 1900 my family, made up of Jewish immigrants from Hungary and their American-born children, lived in Homestead’s Second Ward, along with most of the town’s Eastern European immigrants.[1]  This community dates to the opening of the town’s steel mill in 1881, and for the next two decades the town watched warily as its population of “Huns” and “Slavs” grew.[2]  Their presence felt like a threat; they were perceived as undercutting the wages and taking the jobs of English-speaking men.  Additionally, the immigrants’ social mores and impoverished conditions, especially as manifested by the young, single men who predominated, appeared as rowdiness, drunkenness, and uncleanliness that threatened the larger social order.  Even as the immigrants began to organize family, religious, and social structures that matched those in longer-established American communities, the town’s older residents disdained their new neighbors and held them apart.

This reluctance to integrate manifested itself in the records kept about who lived in Homestead.  The town’s early city directories reflect a clear bias around who was worth counting—largely English-speaking men.  In the 1890s the local men who assembled the directories did not make a serious attempt at including their Eastern European immigrant neighbors.[3]

Homestead stagnated for much of the decade after the strike and the subsequent depression in the steel industry.  Near the turn of the century, when conditions finally picked up, industrialists expanded local operations.  As a result, the town’s population began to soar, with an influx of English speakers and Eastern Europeans alike.

This growth brought new possibilities to Homestead.  One that had long eluded the town was free postal delivery to their homes, instead of having to wait in increasingly long lines at the post office to pick up their letters.  Local officials needed to prove a population greater than 10,000 in order to petition the federal government for this service.[4]

In March of 1900, the town conducted a police census to prove they met that requirement. [5]  It took two days to canvass most of the borough, but in the Second Ward, even though they had started with a larger team of men, they unexpectedly needed a couple extra days.  The newspaper explained,

The canvassers met with several surprises in making their rounds.

In one house of six rooms on Sixth avenue 30 people were found and in one hotel 50 boarders….Nearly all of these men are new comers.[6]

As the canvassers worked, they predicted a count of 4,000 people in the Second Ward.[7]  Instead,

The News-Messenger, March 17, 1900

The result of the census taken in the second ward caused a great deal of surprise.  No one thought for a moment that there were over 6,000 people in the ward and the question is being freely asked, “where are they all kept?”  The ward shows a population of nearly as many as any three of the other wards combined, yet in area it is not as large as some of the others.[8]

The newspaper went on to marvel at the features of life in the ward—the dense concentration of homes, the large number of boarders and children.  Many of the homes were arranged in “courts,” groups of shabby, poorly maintained houses, with no running water and little sunlight or ventilation, encircling trash-filled courtyards where children played amongst filthy outhouses. The courts were a public health nightmare, the result of the terrible wages the steel mill paid its immigrant laborers.  The canvassers had seen for themselves the bad living conditions of their poorest neighbors, but these discoveries the newspaper concealed.[9]

It was a time for celebration: with 15,006 residents (40% from the Second Ward), Homestead exulted in its “wonderful growth” from “a little village in 1880” to “the largest borough in the state.” A committee of politicians went to D. C. a few days later with the police census results in hand.  Free postal delivery for Homestead started in July.[10]

Meanwhile, on June 1, the Federal Census began.  The two men assigned to canvass the two enumeration districts of the Second Ward were Stephen Terski, a 21 year-old mill laborer, and Anthony Karabasz, a 29 year-old druggist.  Both were native Polish speakers, which made them the right fit for at least some of the ward’s immigrants, but they were also non-residents of Homestead, which made them unfamiliar with the unique characteristics of the ward.  The overcrowded courts had challenged the very men who patrolled the neighborhood daily.  How would two outsiders fare?

After nine days of work, Anthony submitted 35 pages.  Stephen, however, started three days late, and after two weeks, submitted 38 pages.  It then appears that Stephen was one of the county enumerators called back to work to make corrections, because a week later he added another page.[11]

When the results of their work came out in mid-July, their numbers fell short of the figures from the police census.[12]

Ward Police census
(March 1900)
Preliminary federal census
(July 1900)
First 2,376 2,115
Second 6,062 4,286
Third 2,196 2,191
Fourth 2,323 2,362
Fifth 2,049 1,957
Total 15,006 12,314*

* I know these numbers don’t sum correctly, but this is what was printed in the paper.

A great deal of complaint is being made about the manner in which the census of the second ward was made…It is said that in many cases family after family was missed where more than one family occupied a single house, the numerator (sic) not being familiar with the neighborhood.  But this is not all.  The claim is also made that blocks of houses were missed and one case has been brought to our notice where a boarding house containing 28 people was not visited at all, and the inmates are all English speaking people at that.[13]

Nevertheless, a week later the county supervisor was one of the first census supervisors in the nation to send his returns onto the Census Bureau in D. C.[14]  The Pittsburgh papers crowed over his “executive ability, as he had a much harder field to cover than some of the others,” but Homestead was not willing to let the matter rest.

Last week [the county census supervisor] sent for the cards used in making the police census last March and compared them with those handed in by his men, and found that they tallied all right in all wards but the Second, and here he found a great discrepancy, there being a difference of about 1,700 in the figures.  Name after name appeared on the police census cards that were omitted from those taken by the census enumerator.

The News-Messenger, August 23, 1900

Initially the county supervisor was “not willing to admit that there was anything wrong with the census taken by his men.”  He claimed that the 12.5% reduction in the steel mill’s workforce in the spring disproportionately affected the foreigners of the Second Ward, who must have moved elsewhere in search of work.  Eventually, though, he gave in and “ordered a revise (sic) to be taken…in order to set at rest all doubts about the matter, the same man being employed as made the first canvass.”[15]

Cross-outs of duplicate people (click to enlarge).

Strangely, the Homestead newspaper did not report on the second canvass, nor its results.  The only clue about what happened is four undated pages tacked to the end of Stephen’s returns, which jump from street-to-street with only a few households on each, different handwriting for each section, and cross-outs of people who turned out to be duplicates.  These pages seem to reflect some sort of attempt to catch missing people, although it is unclear whether they date to the June “correction” or the August “re-canvass.” Either way, they covered nowhere near 1,700 missing people, which would have required 34 pages.

Despite whatever happened during the re-canvass, the final count for the Second Ward dropped well below the preliminary count!

Ward Police census
(March 1900)
Preliminary federal census
(July 1900)
Final federal census [16]
First 2,376 2,115 2,259
Second 6,062 4,286 3,832
Third 2,196 2,191 2,184
Fourth 2,323 2,362 2,312
Fifth 2,049 1,957 1,967
Total 15,006 12,314* 12,554

After all that, Homestead was still stuck with a lower count.  The federal census, not the police census, was the one that mattered.

The police census has been lost to time.  Without it, is it possible to settle the issue for once and for all about the quality of the 1900 federal census of Homestead?  Was Homestead right that a large number of people who should have been counted were not?  Or was the county census supervisor right that all the missing people had actually left town?

As it turns out, Homestead’s residents were enumerated for a third time in the same four-month period of 1900.  On the very day the police census results were published, the town’s newspaper announced that a new directory was in the works.  This time, however, “An endeavor was made this year to obtain the names of the principal part of the foreign population, a feature largely omitted heretofore in publishing a Directory in this place.”  While Homestead had a long way to go to accept its immigrant residents—arguably, it would take another generation for their American-born, English-speaking children to break through the prejudice—the police census at least forced the town to acknowledge that that their immigrant residents were not such a fringe minority after all.[17]

The directory did not aim for the comprehensiveness of a census.  By design, it included only a fraction of all residents, mostly employed, adult men, along with the few women who worked outside the home and the smattering of students at the high school level or beyond.  But it is possible to compare the Jewish residents in the 1900 directory to those in the 1900 federal census, in order to assess the overall quality of the federal census.

Homestead’s 1900 city directory lists 32 Jewish people in the Second Ward, 13 of whom did not make it into the census—an undercount of 41%.[18]  The data in the above table shows that the total federal census undercounted the Second Ward by 37% compared to the police census.  The shortfall in Jewish residents aligns with the overall undercount in the federal census.  The federal census missed more than a third of the ward’s residents.

By comparison, for Jewish people outside of the Second Ward, the federal census undercount was 1 out of 14, or 7%.  This lower undercount fits with the closer match between the police and federal census counts for the other wards.

Using many other kinds of records to document the presence of Homestead’s Jewish residents, I estimate that there were about 200 Jews in the Second Ward at the time the 1900 census was taken.  Perhaps not much ought to be extrapolated from just this one, small group.  Their undercount, however, cannot be explained by steel worker layoffs, since none of the men were steel workers.  Some of the missing people include my own family and leading Jewish shopkeepers like Benjamin Friedlander (called “popular” by the newspaper in 1900), Morris and Meyer Grinberg (“well known”), and Abraham Skirball (“prominent”)—people I am absolutely certain lived in Homestead before, during, and after all these enumerations.[19]  Actually, one might argue that Homestead’s Jews should have been missed less often than their neighbors, not only because their stores brought them to the town’s attention, but also because fewer lived in crowded courts.

Representative newspapers ads from this period for Friedlander, Skirball, and the Grinbergs (click to browse).

Despite the limitations of my analysis, I believe it confirms what Homesteaders knew in the summer of 1900:  the census of their immigrant district was no good.  It seems clear that all of the challenges of counting the residents of the ward, from their crowded living conditions and diversity of languages, to their mistrust of outsiders (remember that this was the first census since the strike), yielded an enumeration that is demonstrably deficient.

Ultimately, other than a theoretical loss of prestige in its competition with neighboring towns, it is unlikely that the difference between 12,500 and 15,000 people affected Homestead at the time.  Today, however, no one who uses the Homestead 1900 census knows that it is marred by such serious omissions.  And yet, this full story, resurrected here after 122 years, leads one to wonder how many such issues—both those that were once known, and those that were never revealed—affect our ability to research our own ancestors and to understand the true context of the places in which they lived.


This post is part one in my series exploring enumeration issues in Homestead censuses! Keep on reading:

  1. “Where are they all kept?”: Undercounted Immigrants in the 1900 Census
  2. When Henry Silverstein Got Cold: Fraud in the 1920 Census
  3. The One Where I Obsess Over All the People I Can’t Find in the Census

Thank you to Renée Carl and Lina Goldberg for reviewing a draft of this essay.


[1] The Second Ward (often called just “the ward”) was a small area between the river and Eighth Avenue and City Farm Lane and McClure.  Most of it was demolished in 1941 to expand the town’s steel mill.

[2] The immigrants were primarily comprised of Slovaks and included Hungarians, Poles, Lithuanians, and others.

[3] The News-Messenger, June 28, 1900, p. 1.  It’s pretty easy to confirm this omission, though, just by reviewing the names of all the people with profession “laborer” or with addresses on Second Ward streets. Or, compare the 1891 or 1892 directories to the list of laborers whom Frick named in his Congressional testimony after the strike (the only surviving employment list from the early years).  This omission, coupled with the loss of the 1890 census due to fire, makes the deficiencies of the 1890s Homestead directories especially challenging for researchers.

[4] They also had to prove more than $8,000 in post-office receipts.  (Somehow this article is the second time I am writing about Homestead’s free delivery movement.  The first was in the story of the second Friedlander envelope.)

[5] A census conducted by the police officers themselves under the supervision of their chief, although it appears that business men temporarily sworn in as police officers to help out!  (The News-Messenger, March 14, 1900, p. 1.)

[6] The News-Messenger, March 15, 1900, p. 1.

[7] The News-Messenger, March 16, 1900, p. 1.

[8] The News-Messenger, March 17, 1900, p. 4.

[9] Most details from Homestead: The Households of a Mill Town, Chapter 9.  Not until 1909, when the book’s findings became public, was the town forced to acknowledge publicly how their poorest residents lived.  Homestead’s newspaper admitted that the report amounted to a “black eye,” but protested,

It is true that there are many undesirable conditions here, but they are no worse than those to be found in any other manufacturing town…These professional writers are sent out to get the dark side of life and that is all they look for…In this way Homestead had been shamefully misrepresented and the impression has gotten broadcast that this is a community made up solely of illiterate foreigners, who live little cattle, when in fact, the greater part of the population is made up of the better class and the town can boast of more literary clubs and societies than any town of its size in the country (The Daily Messenger, January 5, 1909, p. 4).

[10] The News-Messenger, March 17, 1900, p. 4 and March 19, p. 1. Waiting until July was an unsatisfactory result, but the town had to wait for new Congressional appropriations.

[11] Pittsburgh Commercial Gazette, June 28, 1900, p. 7.  Stephen took District 400, the area below Fifth Avenue.  Anthony took District 401, the area above Fifth Avenue.  The population density was far greater below the tracks in District 400.

[12] The News-Messenger, March 17 and July 20, 1900.

[13] The News-Messenger, July 24, 1900, p. 4.

[14] Pittsburgh Commercial Gazette, July 31, 1900, p. 4.

[15] The News-Messenger, August 23, 1900, p. 1.

[16] Final Federal numbers from 1900 Census: Volume I. Population, Part 1, p. 331, 472.

[17] The News-Messenger, March 17, 1900, p. 3; June 28, 1900, p. 1.

[18] The 32 Jews in the Second Ward includes some people for whom I have no other records hinting where they lived, so I’m erring on the side of the ward since that was where Jews concentrated in Homestead at that time.

[19] Newspaper quotes from April 21, 1900, February 8, 1900, and Skirball obituary on June 2, 1905.

The census miss rate may actually be higher.  I found 12 heads of household in neither the census nor the city directory whom other evidence suggests lived there then.  Eight of those I feel certain were there then; all were Second Ward residents, half of whom lived in households the census missed.

For what it’s worth, the census included 10 Jewish people who were not in the city directory, although most of their absences have likely reasons.  That said, given the full set of Jewish people I believe to have been in Homestead at this time, the city directory’s miss rate is far lower than the census’.

  2 comments for ““Where are they all kept?”: Undercounted Immigrants in the 1900 Census

  1. Theresa
    March 22, 2022 at 3:45 pm

    I am in awe of your research skills. Kudos!

  2. March 29, 2022 at 9:47 am

    This is extraordinarily helpful; I’m grateful to you for your research and sharing of it, and that I stumbled across it! I have some ancestors who I believe lived in Homestead for a few years, right around the turn of the century, but I don’t have very reliable evidence and have not been able to find them in the 1900 census. The couple were both immigrants, Russian Polish and German Polish, with children born in Pennsylvania beginning around 1895. I’ll be checking out the rest of your website and re-reading this post to better apply it’s revelations to my own research. Here’s a link to my blog posts about this family line, in case you’re interested: https://lieselsfamilyhistoryfinds.blogspot.com/2021/07/frank-wisniewski-and-mary-novak-series.html

    Thanks again!

Leave a Reply

Your email address will not be published. Required fields are marked *