Searching the Census Without Using the Soundex

Robert W. Marlin guides us through this tricky problem.

In order to understand “Searching The Census Without Using the Soundex”, it is important that you have some understanding of the tools available to you. The US Population Censuses are one of the finest set of documents available to genealogical researchers. These censuses have been take every 10 years since 1790. At the present time, almost all census schedules are available to the public on microfilm, up to the year 1920. The only exception is the 1890 census, the majority of which was destroyed in a fire. Early in the year 2003, the 1930 census will also become available to the public.

There are actually two parts to most of these censuses. They consist of the actual censuses schedules themselves and the indexes which help to give you access to the schedules. Many of the earlier censuses from 1790 to 1880 are already indexed in book form. Most of this work has been done by sources other than the National Archives. Finding your ancestors in these early censuses is not always easy or simple. The censuses from 1790 to 1840 only listed the head of a household. The 1850 census was the first to list each family member, their age and their occupation.

Enumeration districts of a large city (Seattle, Washington).

Soundex Indexes on film exist for the 1880, 1900, 1910 and 1920 censuses. Currently, the National Archives publishes four handbooks which outline a lot of information about how each individual census was done. The first is entitled Federal Population Censuses 1790-1890. The second, third, and fourth handbooks are for the 1900, 1910 and 1920 censuses respectively. It is a good idea to browse through them as they have a lot of information which could help with your research. The 1880 census was the first one to have a Soundex Index. However, the way it was created is somewhat flawed. This index only includes households with children under 10 years of age, a fact that many researchers don’t seem to be aware of. The 1910 Soundex Index is only available for about 25 selected states. I have talked with researchers who are not aware that the 1910 census has any index.

The Soundex indexes now available from the National Archives were not created until the Depression of the 1930s. They were created primarily to create jobs for thousands of destitute Americans. All of the entrees were transcribed by hand onto file cards which were later microfilmed. They were filed according to what is known as the Soundex Code. This special coding puts all names which sound alike into the same code number. A name such as Cooper might be spelled Cooper, Couper, Coupper or even Coopar. All of these entries would be indexed in the same group. The people who did this work were not necessarily educated and probably few had any interest in genealogy. They were starving people looking for any way to feed their families. These indexes have many omissions and errors, but without them we would be hard pressed for any degree of success with our pursuit.

In the genealogy department on the fourth floor of the Orlando Public Library are a series of black steel filing cabinets. In floor space they only occupy about 64 square feet. There are about 16 of them with 11 drawers each. One of the cabinets has three empty drawers. The other 173 drawers are filled with rolls of microfilm. Each drawer contains 75 rolls of microfilm which places the total number of microfilm rolls at 12,975. They contain every census schedule from 1790 to 1920. Every bit of census information that it is possible for me to uncover about my ancestors is contained in that group of cabinets. That’s the good news.

Enumeration districts of a small town in West Virginia.

The bad news is that the Soundex Index films which normally provide quick access to these schedules are not available at this library. The only place I know of where both the schedules and the indexes are both available at the same time is at the National Archives or at its regional centers across the country. Most libraries simply cannot afford the expense of both sets of film.

For some years, in order to use the schedule entries, I had to first order the index film at my local Family History Center, wait two weeks for the film to arrive, look up the people I was looking for and then go back to the library to look at the schedule. This seemed like a waste of a lot of time and I began to look for some way of bypassing the use of the index. My method was developed by a lot of trial and error, but has now reached a point where I am able to find what I want in most censuses without using an index. In order for me to get started it was necessary to do some census history homework.

The US is primarily broken down geographically into state and county levels. The exceptions are Louisiana, Delaware and Connecticut. Louisiana has parishes instead of counties, while Connecticut and Delaware have townships. The county level is broken down into a number smaller units. Most counties were subdivided into townships. The township can contain cities, villages, hamlets and boroughs. Remember, these are geographical subdivisions which are not directly related to the census. However, they are extremely important because they are recorded in the census schedules.

When the original schedules were microfilmed, in many cases, the townships in a county were placed in alphabetical order. A fair sized city was sometimes placed in front of the township it was in. The villages or small towns were sometimes placed somewhere within the township. The top of the schedule page usually indicated where the village entries started and ended. There were also two census related subdivisions. These were the Supervisor’s District and the Enumeration Districts. The Enumeration District or E.D. is of special use in locating streets addresses in the censuses. There also exists on microfiche a special index list of E.D.s in 39 cities throughout the US.

This brings us to an additional set of index films which are called “Descriptions of Enumeration Districts”. They are extremely helpful when using the 1880, 1900, 1910 and 1920 censuses. This is the one we will be the most interested in using in order to tie the geographical location to an E.D. With some exceptions, many of these E.D.s follow each other numerically and geographically. However, the E.D. numbers as well as the size and shape of the E.D., changed from census to census. You cannot use the E.D. number found in the 1910 census to locate someone at that same address in the 1920 census. It might be wise to keep one thing in mind. You can be quite certain that the people who designed the methods of conducting and recording the censuses had no idea that many years later we would be searching the schedules for our ancestors. Their only concern was getting the information in the quickest most efficient manner for them. Their method was not for the benefit of future genealogists. In spite of multitudes of errors and omissions, I’m thankful for the great job they did.

Enumeration Districts Census schedules are arranged by state or territory and then by county, and beginning in 1880 by enumeration district (E.D.). E.D.s were the areas that an enumerator covered in taking the census. To consult the schedules for a particular town, a minor civil division or geographical area, or a ward of a large city, one must know the enumeration district. National Archives Microfilm Publication T1224 describes E.D. boundaries as they were in 1920; present-day boundaries may not be the same. Rolls 41-60 of Descriptions of Census Enumeration Districts 1830-1890 and 1910-1950 (T1224) identify the enumeration district number assigned within the state, county, and city for the 1920 census. The descriptions are arranged alphabetically by state and thereunder by supervisor’s district, which is a large geographic area that covers several counties. The descnptions are then arranged by county and thereunder by township or city. The E.D. is listed on the left-hand side of the page. (See example above.) Quick Guide to Finding Ancestors Without Soundex 1. From a street address in a city, or town name in a rural community, look up the enumeration district in which the street or town appears. 2. To find the E.D. number for the street consult the Descriptions of Enumeration Districts indexes available at state libraries and the LDS Family History Library. There is also a microfiche index of E.D.s in 39 cities in the US. 3. Having found the E.D. in which the street or rural town appears consult the Census Handbook Index to see which film number this E.D. is on. 4. Obtain film for this E.D. and search the street or rural town in which your ancestor lives. 5. This census search system only works with the 1880, 1900, 1910 and 1920 censuses since these censuses list the street name and house number on each page of the schedule. 6. Practice is needed to be comfortable with this system but since there are so many errors and omissions in the Soundex extractions this method will provide an alternative to locating ancestors when their street or rural town location is known.

The 1880, 1900,1910 and 1920 census schedules have one thing in common. Each page of the schedule contains both the street name and the house number. This is part of what makes my method work. The system is simple. I’m going to give you actual examples of how I have used it to find ancestors, even when I couldn’t find them in the Soundex Index. I’ll give you an example of the use in a large city like New York City in 1910 and then I’ll show you how I found a family in rural Georgia in the 1920 Census.

When trying to find census schedules for ancestors who lived in large cities, it is quite important to know the exact street address of the family you are looking for. Even without the exact address you can still scan a neighborhood, but this can be time consuming. The address can be obtained from birth certificates, death records, marriage licenses or city directories. Then you must be able to locate the streets which border that address. The procedure is somewhat like plotting coordinates on a map. My father’s birth certificate indicated that he was born on February 10, 1910 at 287 Henry Street, New York City. (When I was doing this research the E.D. index for New York City was not available to me.) If they were living at this address when the census was taken, all I had to do was find that address in the census. Using the street address of 287 Henry Street as a focal point, I then consulted several modern New York City Street Maps. They showed that Henry Street was only seven or eight blocks long. It ran between East Broadway and Madison Street. The street number of 287 was located between Scammel and Grand Streets. This breakdown of house numbers can be found in most cross-reference telephone directories. This is where I first used the E.D. Description microfilm. It took very little effort to find the film roll for New York City. Scanning through the E.D. description I quickly located several streets in the group listed above. Then it was simply a question of looking up the schedules for that E.D. This was done by consulting the 1910 Census Handbook described earlier. For each film roll it lists the E.D.s the film contains. Within a few minutes I was scanning though the schedule film roll. It took about 15 minutes before I found Henry Street and quickly located my father’s family in the 1910 census. Was this luck or due diligence? My earliest efforts in developing my method involved a lot of trial and error.

My methods are now a little more sophisticated than what I just described. For example, I use an inexpensive software program called Streets-USA. It is available on CD-ROM for less than $10. You can bring up a city map and then key in an intersection or exact street address. The program brings up a map which shows all the streets in the area you are looking for. In any case, that is how I located someone the first time I tried using the census with Soundex. This method described can be used anywhere. I have also used it for places like Buffalo, New York and Philadelphia and Los Angeles.

The second example works even better for rural areas. Several years ago, one of my cousins had tried unsuccessfully to do research on his father. He had little information to go on. He only knew that his father had died in 1935 and that that he was from Henderson, Georgia. This certainly was not much to go on. I simply looked on a map and found that Henderson was located in Houston County. I looked up the film roll number and found that all of Houston County for the 1920 census was on one roll of film. My worst case scenario would be that I would have to search an entire roll of film. After a couple of minutes I realized that the schedule showed towns and villages. Within 10 minutes I found his family in Henderson. It was so easy that further research showed that most rural areas were also divided the same way. I had occasion recently to do some research in Ansonia, Ohio for one of my clients. The Darke County microfilm was neatly alphabetized by township. The town of Ansonia was in Brown Township, right where it belonged. I did not have a street address so I had to scan the street names and numbers on the left hand side of each page. It took about 15 minutes before I found what I was looking for.

This methodology is not an exact science. It is only a guideline which works for me and I hope it will work for you. I have not had any occasion in the past five years to refer to Soundex Indexes. In that time I have located dozens of schedules. If you discover any additional methodology which you feel would improve this technique, please let me know.

This article originally appeared in the May/June 1999 issue of Family Chronicle.